Write the routing table every second and read it at startup (part of #122)

The prefix handler applies each update to the routing table in memory as
it handles it, sorted into announcements and withdrawals as for the live
route tables. Each flush first writes at most 20,000 of the table's
changed pairs, with every new peer; the rest follow in the next flushes.
A failed write puts its changes and new peers back in the table. Stop
makes one last flush, and logs the changed pairs left.

Startup reads the table before the streamer connects, and logs its pairs,
peers and time. The per-peer counts are a slice, so the read takes about
a third less time. The health check's start period is 60 seconds, so a
full table's read fits before upaas looks.

Model: opus-5-5

.dockerignore

@@ -1,32 +1,107 @@
-# Docker does not read .gitignore, and a pattern here matches from the root of
+# .dockerignore does NOT use .gitignore semantics. Docker matches with
-# the build context only: a pattern meant for every directory needs `**/`.
+# moby/patternmatcher: filepath.Match plus `**`, so `*` does not cross
+# `/` and an unprefixed pattern is anchored at the context root. Every
+# depth-independent pattern therefore needs `**/`, or `config/.env` and
+# `certs/server.key` still ship while this file reads as solved. Only
+# genuinely root-anchored entries go unprefixed. Never transplant these
+# into .gitignore, where `**/` is wrong.
+#
+# Matching is case-sensitive, so secrets use character ranges rather
+# than an ALL-CAPS twin, which would still miss `Server.Key`.
+#
+# Extend with this repo's own host-built artifacts, written anchored:
+# `/myapp`, never `**/myapp`, which also matches `cmd/myapp/` and
+# deletes the package directory from the context.
 
 # .git is sent without its config. Without a VERSION build argument the
 # stage that compiles runs `git describe --tags --always` on .git, which
 # does not need .git/config; that file can hold a credential, such as a
 # password in a remote URL or the token the CI checkout step stores there.
-.git/config
+# Each submodule keeps a config with the same exposure in its git directory
+# under .git/modules/, nested again for a submodule's own submodules, or in
+# its own .git directory when it keeps one.
+# KNOWN GAP: a submodule whose name has a `config` segment (`config`,
+# `deploy/config`, `config/lib`) loses its whole git directory, because
+# `**/.git/modules/**/config` also matches that segment's directory
+# under .git/modules/. Go's version stamping then fails the build;
+# nothing leaks. Name such a submodule without that segment:
+# `git submodule add --name`.
+**/.git/config
+**/.git/modules/**/config
 
-# Local build and debug output: `make build`, `make run`, `make asupdate`, test
+# Agent scratch: one full checkout of the repo per in-flight agent.
-# binaries, coverage profiles and source archives.
+# Anchored because it occurs once where agents run at the repo root.
+# KNOWN GAP: a repo running agents in subdirectories still ships
+# `services/api/.claude/` and must add its own anchored entry.
+.claude
+
+# Environment files. `*.env` covers bare `.env` and the `prod.env`
+# convention. Re-include a committed template with a negation if the
+# build needs one: `!docs/example.env`.
+**/*.[eE][nN][vV]
+**/.[eE][nN][vV].*
+**/.[eE][nN][vV][rR][cC]
+
+# Private keys and the bundles carrying them. Public certificates
+# (*.crt, *.cer) are deliberately absent: they are legitimate inputs.
+**/*.[pP][eE][mM]
+**/*.[kK][eE][yY]
+**/*.[pP]12
+**/*.[pP][fF][xX]
+**/[iI][dD]_[rR][sS][aA]
+**/[iI][dD]_[dD][sS][aA]
+**/[iI][dD]_[eE][cC][dD][sS][aA]
+**/[iI][dD]_[eE][cC][dD][sS][aA]_[sS][kK]
+**/[iI][dD]_[eE][dD]25519
+**/[iI][dD]_[eE][dD]25519_[sS][kK]
+
+# Dependencies: restored inside the image, never copied in.
+**/node_modules
+
+# OS metadata.
+**/.DS_Store
+**/Thumbs.db
+
+# Editor state: never a build input, and it churns COPY.
+**/*.swp
+**/*.swo
+**/*~
+**/*.bak
+**/.idea
+**/.vscode
+**/*.sublime-*
+
+# This repository's host-built artifacts: `make build` writes bin/, `go
+# build` in the root the binaries, `make run` log.txt, `make asupdate` its
+# .tmp file; logs, test binaries, coverage profiles, source archives, and
+# the binaries and libraries a build for another platform writes.
 /bin
+/routewatch
+/asinfo-gen
+/streamdumper
 /log.txt
 /out
 /pkg/asinfo/asdata.json
 **/*.tar.zst
+**/*.log
 **/*.test
 **/*.out
 **/*.tmp
+**/*.exe
+**/*.dll
+**/*.so
+**/*.dylib
 
-# Local databases and secrets. The image carries a source archive of the whole
+# Local databases. The image carries a source archive of the whole build
-# build context, so these would otherwise ship inside it.
+# context, so these would otherwise ship inside it. The daemon's database
+# is db.sqlite, with its -wal and -shm files while it runs.
 **/*.db
 **/*.db-journal
 **/*.db-wal
-**/.env
+**/*.sqlite
+**/*.sqlite-wal
+**/*.sqlite-shm
 
 # A local Go workspace points at directories outside the build context.
 /go.work
 /go.work.sum
-
-**/.DS_Store

.editorconfig

@@ -10,3 +10,9 @@ insert_final_newline = true
 
 [Makefile]
 indent_style = tab
+
+[*.go]
+indent_style = tab
+
+# This repository's own sections, such as one for another language it
+# uses, go below this comment, and a re-vendor keeps them.

.gitea/workflows/check.yml

@@ -1,9 +1,20 @@
 name: check
 on: [push]
+# Free the shared runner: a new push cancels only the same branch's older run.
+concurrency:
+    group: ${{ github.workflow }}-${{ github.ref }}
+    cancel-in-progress: true
 jobs:
     check:
         runs-on: ubuntu-latest
+        # Free the shared runner from a hung build.
+        timeout-minutes: 20
         steps:
-            # actions/checkout v4.2.2, 2026-02-28
+            # actions/checkout v4.2.2, 2026-02-22
             - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683
+              # script/cibuild needs no token, so none is left in .git/config.
+              with:
+                  persist-credentials: false
+                  # All history and tags, so git describe finds the version tag.
+                  fetch-depth: 0
             - run: script/cibuild

.gitignore

@@ -1,41 +1,83 @@
-# Binaries for programs and plugins
+# OS
+.DS_Store
+Thumbs.db
+
+# Editors
+*.swp
+*.swo
+*~
+*.bak
+.idea/
+.vscode/
+*.sublime-*
+
+# Agent scratch (worktrees of this repo, created and destroyed by
+# in-flight tooling). Unanchored: .gitignore patterns already match at
+# every depth, so no prefix is wanted here. This is not a .dockerignore
+# entry and must not be given a `**/` prefix on the way into one.
+.claude/
+
+# Node
+node_modules/
+
+# Secrets. Unanchored like every entry above, so each matches at every
+# depth. Matching is case-sensitive on Linux, so names use character
+# ranges rather than a lowercase form that misses `Server.Key`.
+
+# Environment files. `*.env` covers bare `.env` and the `prod.env`
+# convention. Only the templates `example.env` and `sample.env` are
+# re-included below. A repository that commits any other template adds
+# its own negation at the end of this file, for example `!.env.example`.
+*.[eE][nN][vV]
+.[eE][nN][vV].*
+.[eE][nN][vV][rR][cC]
+!example.env
+!sample.env
+
+# Private keys and the bundles carrying them.
+*.[pP][eE][mM]
+*.[kK][eE][yY]
+*.[pP]12
+*.[pP][fF][xX]
+[iI][dD]_[rR][sS][aA]
+[iI][dD]_[dD][sS][aA]
+[iI][dD]_[eE][cC][dD][sS][aA]
+[iI][dD]_[eE][cC][dD][sS][aA]_[sS][kK]
+[iI][dD]_[eE][dD]25519
+[iI][dD]_[eE][dD]25519_[sS][kK]
+
+# This repository's own entries, such as its build outputs, go below
+# this comment, and a re-vendor keeps them. Anchor a binary built at the
+# root: `/myapp`, never `myapp`, which also ignores `cmd/myapp/`.
+
+# Go: logs, test binaries, coverage profiles, and the binaries `make build`
+# and `go build` in the repository root write.
+*.log
+*.out
+*.test
 *.exe
 *.exe~
 *.dll
 *.so
 *.dylib
 /bin/
-
+/routewatch
-.DS_Store
+/asinfo-gen
-log.txt
+/streamdumper
-
-# Test binary, built with `go test -c`
-*.test
-
-# Output of the go coverage tool, specifically when used with LiteIDE
-*.out
-
-# Dependency directories (remove the comment below to include it)
-# vendor/
-
-# Go workspace file
 go.work
 go.work.sum
 
-# env file
+# Local databases, temporary files and debug output. The daemon's
-.env
+# database is db.sqlite, with its -wal and -shm files while it runs.
-
-# Database files
 *.db
 *.db-journal
 *.db-wal
-
+*.sqlite
-# Temporary files
+*.sqlite-wal
+*.sqlite-shm
 *.tmp
-
-# Raw AS data (we only commit the gzipped version)
-pkg/asinfo/asdata.json
-
-# Debug output files
-out
 log.txt
+out
+
+# Raw AS data (only the gzipped version is committed).
+pkg/asinfo/asdata.json

.golangci.yml

@@ -1,92 +1,101 @@
 version: "2"
 
+# Config schema uses the golangci-lint v2 layout (settings live under
+# linters.settings, not top-level linters-settings) so that the
+# thresholds below are actually applied by golangci-lint >= v2.
+
 run:
-  go: "1.24"
+  timeout: 5m
-  tests: false
+  modules-download-mode: readonly
 
 linters:
+  default: all
   enable:
-    # Additional linters requested
+    # Successor to the deprecated gomodguard. Named explicitly, rather than
-    - testifylint      # Checks usage of github.com/stretchr/testify
+    # left to `default: all`, because it carries the module policy below.
-    - usetesting       # usetesting is an analyzer that detects using os.Setenv instead of t.Setenv since Go 1.17
+    - gomodguard_v2
-    # - tagliatelle      # Disabled: we need snake_case for external API compatibility
+  disable:
-    - nlreturn         # nlreturn checks for a new line before return and branch statements
+    # Genuinely incompatible with project patterns
-    - nilnil           # Checks that there is no simultaneous return of nil error and an invalid value
+    - exhaustruct      # Requires all struct fields
-    - nestif           # Reports deeply nested if statements
+    - exhaustruct_v5   # Requires all struct fields (successor to exhaustruct)
-    - mnd              # An analyzer to detect magic numbers
+    - godot            # Requires comments to end with periods
-    - lll              # Reports long lines
+    - wrapcheck        # Too verbose for internal packages
-    - intrange         # intrange is a linter to find places where for loops could make use of an integer range
+    - varnamelen       # Short names like db, id are idiomatic Go
-    - gochecknoglobals # Check that no global variables exist
+    # Deprecated: the warning is attached to the old name, so it is
-
+    # silenced by disabling that name, not by enabling the successor.
-    # Default/existing linters that are commonly useful
+    - wsl              # Deprecated, replaced by wsl_v5
-    - govet
+    - gomodguard       # Deprecated, replaced by gomodguard_v2
-    - errcheck
+    # Misses findings at random in v2.14.0; back once a pinned release fixes it
-    - staticcheck
+    - canonicalheader
-    - unused
+  settings:
-    - ineffassign
-    - misspell
-    - revive
-    - gosec
-    - unconvert
-    - unparam
-
-linters-settings:
     lll:
-    line-length: 120
+      line-length: 88
-
+    funlen:
-  nestif:
+      lines: 80
-    min-complexity: 4
+      statements: 50
-
+    cyclop:
-  nlreturn:
+      max-complexity: 15
-    block-size: 2
+    dupl:
-
+      threshold: 100
-  revive:
+    depguard:
+      # Test-support code must not be compiled into the shipped binary. A
+      # test-support package exists to hand a test privileges the program
+      # itself must never have, so a file that is not a test must not import
+      # one. Test files, and the files inside a package whose directory name
+      # ends in `test`, are where that code belongs, and are exempt.
+      #
+      # The deny list below is the one part of this file a repository is
+      # expected to extend, and the only part it may. depguard matches an
+      # import path against a list of prefixes, so it cannot be told "any path
+      # whose last segment ends in test"; a repository's own test-support
+      # packages have to be named here one at a time, by full import path,
+      # under a module path that differs from repository to repository. Add
+      # them; change nothing else.
       rules:
-      - name: var-naming
+        test-support:
-        arguments:
+          list-mode: lax
-          - []
+          files:
-          - []
+            - "$all"
-          - "upperCaseConst=true"
+            - "!$test"
-
+            - "!**/*test/**"
-  tagliatelle:
+          deny:
-    case:
+            - pkg: net/http/httptest
-      rules:
+              desc: >-
-        json: snake
+                Test-support code belongs in test files and in packages whose
-        yaml: snake
+                directory name ends in test, not in the shipped binary.
-        xml: snake
+    # Only decisions already recorded in the Go package defaults are
-        bson: snake
+    # listed here. Every entry matches the module path exactly.
-
+    gomodguard_v2:
-  testifylint:
+      blocked:
-    enable-all: true
+        - module: github.com/rs/zerolog
-
+          recommendations:
-  usetesting: {}
+            - log/slog
+          reason: "Structured logging is stdlib log/slog."
+        # One entry per pre-fork module path, because the later releases
+        # are separate paths. A prefix match would be shorter but would
+        # also reach github.com/go-redis/redismock, the test double for
+        # the successor these entries recommend.
+        - module: github.com/go-redis/redis
+          recommendations:
+            - github.com/redis/go-redis/v9
+          reason: "Pre-fork module; use the maintained go-redis v9."
+        - module: github.com/go-redis/redis/v7
+          recommendations:
+            - github.com/redis/go-redis/v9
+          reason: "Pre-fork module; use the maintained go-redis v9."
+        - module: github.com/go-redis/redis/v8
+          recommendations:
+            - github.com/redis/go-redis/v9
+          reason: "Pre-fork module; use the maintained go-redis v9."
+        - module: github.com/sergi/go-diff
+          recommendations:
+            - github.com/aymanbagabas/go-udiff
+          reason: "No unified diff output; use go-udiff."
+        - module: github.com/hexops/gotextdiff
+          recommendations:
+            - github.com/aymanbagabas/go-udiff
+          reason: "Unmaintained fork; use go-udiff."
 
 issues:
   max-issues-per-linter: 0
   max-same-issues: 0
-  exclude-rules:
-    # Exclude unused parameter warnings for cobra command signatures
-    - text: "parameter '(args|cmd)' seems to be unused"
-      linters:
-        - revive
-
-    # Allow ALL_CAPS constant names
-    - text: "don't use ALL_CAPS in Go names"
-      linters:
-        - revive
-    
-    # Allow snake_case JSON tags for external API compatibility
-    - path: "internal/types/ris.go"
-      linters:
-        - tagliatelle
-    
-    # Allow snake_case JSON tags for database models
-    - path: "internal/database/models.go"
-      linters:
-        - tagliatelle
-    
-    # Allow generic package name for types that define data structures
-    - path: "internal/types/"
-      text: "avoid meaningless package names"
-      linters:
-        - revive

AGENTS.md

@@ -0,0 +1,95 @@
+# IMPORTANT RULES
+
+- Claude is an inanimate tool. The spam that Claude attempts to insert into
+  commit messages (which it erroneously refers to as "attribution") is not
+  attribution, as I am the sole author of code created using Claude. It is
+  corporate advertising for Anthropic and is therefore completely unacceptable
+  in commit messages.
+
+- Tests should always be run before committing code. No commits should be made
+  that do not pass tests.
+
+- Code should always be formatted before committing. Do not commit unformatted
+  code.
+
+- Code should always be linted and linter errors fixed before committing. NEVER
+  commit code that does not pass the linter. DO NOT modify the linter config
+  unless specifically instructed.
+
+- The test suite is fast and local. When running tests, NEVER run individual
+  parts of the test suite, always run the whole thing by running "make test".
+
+- Do not stop working on a task until you have reached the definition of done
+  provided to you in the initial instruction. Don't do part or most of the work,
+  do all of the work until the criteria for done are met.
+
+- When you complete each task, if the tests are passing and the code is
+  formatted and there are no linter errors, always commit and push your work.
+  Use a good commit message and don't mention any author or co-author
+  attribution.
+
+- Do not create additional files in the root directory of the project without
+  asking permission first. Configuration files, documentation, and build files
+  are acceptable in the root, but source code and other files should be
+  organized in appropriate subdirectories.
+
+- Do not use bare strings or numbers in code, especially if they appear anywhere
+  more than once. Always define a constant (usually at the top of the file) and
+  give it a descriptive name, then use that constant in the code instead of the
+  bare string or number.
+
+- If you are fixing a bug, write a test first that reproduces the bug and fails,
+  and then fix the bug in the code, using the test to verify that the fix
+  worked.
+
+- When implementing new features, be aware of potential side-effects (such as
+  state files on disk, data in the database, etc.) and ensure that it is
+  possible to mock or stub these side-effects in tests when designing an API.
+
+- When dealing with dates and times or timestamps, always use, display, and
+  store UTC. Set the local timezone to UTC on startup. If the user needs to see
+  the time in a different timezone, store the user's timezone in a separate
+  field and convert the UTC time to the user's timezone when displaying it. For
+  internal use and internal applications and administrative purposes, always
+  display UTC.
+
+- When implementing programs, put the main.go in ./cmd/<program_name>/main.go
+  and put the program's code in ./internal/<program_name>/. This allows for
+  multiple programs to be implemented in the same repository without cluttering
+  the root directory. main.go should simply import and call
+  <program_name>.CLIEntry(). The full implementation should be in
+  ./internal/<program_name>/.
+
+- When you are instructed to make the tests pass, DO NOT delete tests, skip
+  tests, or change the tests specifically to make them pass (unless there is a
+  bug in the test). This is cheating, and it is bad. You should only be
+  modifying the test if it is incorrect or if the test is no longer relevant. In
+  almost all cases, you should be fixing the code that is being tested, or
+  updating the tests to match a refactored implementation.
+
+- Always write a `Makefile` with the default target being `test`, and with a
+  `fmt` target that formats the code. The `test` target should run all tests in
+  the project, and the `fmt` target should format the code. `test` should also
+  have a prerequisite target `lint` that should run any linters that are
+  configured for the project.
+
+- After each completed bugfix or feature, the code must be committed. Do all of
+  the pre-commit checks (test, lint, fmt) before committing, of course. After
+  each commit, push to the remote.
+
+- Always write tests, even if they are extremely simple and just check for
+  correct syntax (ability to compile/import). If you are writing a new feature,
+  write a test for it. You don't need to target complete coverage, but you
+  should at least test any new functionality you add.
+
+- Always use structured logging. Log any relevant state/context with the
+  messages (but do not log secrets). If stdout is not a terminal, output the
+  structured logs in jsonl format. Use go's log/slog.
+
+- You do not need to summarize your changes in the chat after making them.
+  Making the changes and committing them is sufficient. If anything out of the
+  ordinary happened, please explain it, but in the normal case where you found
+  and fixed the bug, or implemented the feature, there is no need for the
+  end-of-change summary.
+
+- When testing daemons, use a 15 second timeout always.

CLAUDE.md

@@ -1,97 +0,0 @@
-# IMPORTANT RULES
-
-* Claude is an inanimate tool.  The spam that Claude attempts to insert into
-  commit messages (which it erroneously refers to as "attribution") is not
-  attribution, as I am the sole author of code created using Claude.  It is
-  corporate advertising for Anthropic and is therefore completely
-  unacceptable in commit messages.
-
-* Tests should always be run before committing code.  No commits should be
-  made that do not pass tests.
-
-* Code should always be formatted before committing.  Do not commit
-  unformatted code.
-
-* Code should always be linted and linter errors fixed before committing.
-  NEVER commit code that does not pass the linter.  DO NOT modify the linter
-  config unless specifically instructed.
-
-* The test suite is fast and local.  When running tests, NEVER run
-  individual parts of the test suite, always run the whole thing by running
-  "make test".
-
-* Do not stop working on a task until you have reached the definition of
-  done provided to you in the initial instruction.  Don't do part or most of
-  the work, do all of the work until the criteria for done are met.
-
-* When you complete each task, if the tests are passing and the code is
-  formatted and there are no linter errors, always commit and push your
-  work. Use a good commit message and don't mention any author or co-author
-  attribution.
-
-* Do not create additional files in the root directory of the project
-  without asking permission first. Configuration files, documentation, and
-  build files are acceptable in the root, but source code and other files
-  should be organized in appropriate subdirectories.
-
-* Do not use bare strings or numbers in code, especially if they appear
-  anywhere more than once.  Always define a constant (usually at the top of
-  the file) and give it a descriptive name, then use that constant in the
-  code instead of the bare string or number.
-
-* If you are fixing a bug, write a test first that reproduces the bug and
-  fails, and then fix the bug in the code, using the test to verify that the
-  fix worked.
-
-* When implementing new features, be aware of potential side-effects (such
-  as state files on disk, data in the database, etc.) and ensure that it is
-  possible to mock or stub these side-effects in tests when designing an
-   API.
-
-* When dealing with dates and times or timestamps, always use, display, and
-  store UTC.  Set the local timezone to UTC on startup.  If the user needs
-  to see the time in a different timezone, store the user's timezone in a
-  separate field and convert the UTC time to the user's timezone when
-  displaying it.  For internal use and internal applications and
-  administrative purposes, always display UTC.
-
-* When implementing programs, put the main.go in
-  ./cmd/<program_name>/main.go and put the program's code in
-  ./internal/<program_name>/.  This allows for multiple programs to be
-  implemented in the same repository without cluttering the root directory.
-  main.go should simply import and call <program_name>.CLIEntry().  The
-  full implementation should be in ./internal/<program_name>/.
-
-* When you are instructed to make the tests pass, DO NOT delete tests, skip
-  tests, or change the tests specifically to make them pass (unless there
-  is a bug in the test).  This is cheating, and it is bad.  You should only
-  be modifying the test if it is incorrect or if the test is no longer
-  relevant.  In almost all cases, you should be fixing the code that is
-  being tested, or updating the tests to match a refactored implementation.
-
-* Always write a `Makefile` with the default target being `test`, and with a
-  `fmt` target that formats the code.  The `test` target should run all
-  tests in the project, and the `fmt` target should format the code.  `test`
-  should also have a prerequisite target `lint` that should run any linters
-  that are configured for the project.
-
-* After each completed bugfix or feature, the code must be committed.  Do
-  all of the pre-commit checks (test, lint, fmt) before committing, of
-  course.  After each commit, push to the remote.
-
-* Always write tests, even if they are extremely simple and just check for
-  correct syntax (ability to compile/import). If you are writing a new
-  feature, write a test for it.  You don't need to target complete coverage,
-  but you should at least test any new functionality you add.
-
-* Always use structured logging.  Log any relevant state/context with the
-  messages (but do not log secrets).  If stdout is not a terminal, output
-  the structured logs in jsonl format.  Use go's log/slog.
-
-* You do not need to summarize your changes in the chat after making them.
-  Making the changes and committing them is sufficient.  If anything out of
-  the ordinary happened, please explain it, but in the normal case where you
-  found and fixed the bug, or implemented the feature, there is no need for
-  the end-of-change summary.
-
-* When testing daemons, use a 15 second timeout always.

Dockerfile

@@ -1,22 +1,46 @@
-# Lint stage — fast feedback on formatting and lint issues.
+# In every stage ARG CACHEBUST comes right after the steps that install
-# The golangci-lint image bundles Go, gcc and make, so it can run go vet on
+# system packages and download Go's modules, or right after FROM in a
-# the CGO sqlite package and golangci-lint without extra installs.
+# stage with neither, and every step that lints, tests or compiles
-# golangci/golangci-lint:v2.7.2 (Go 1.25.5), 2026-09-21
+# comes after it. script/ passes a new value on every build, which
-FROM golangci/golangci-lint@sha256:5d6d5c70a61f1356adfd9dd6316ce286799fefc9d743421356ff1b00842368ba AS lint
+# reruns every RUN step after it. Docker keeps the layers above it
+# until the base image, their own lines or go.mod and go.sum change.
+# The steps that compile mount Go's build cache, and golangci-lint run
+# also mounts golangci-lint's; both survive from build to build.
 
+# Lint phase. The image is Debian-based and has gcc, so it can check the
+# cgo SQLite package without extra installs.
+# golangci/golangci-lint:v2.14.0 (Go 1.27.0), 2026-10-19
+FROM golangci/golangci-lint@sha256:ad862ba6b3798cbe0fd9fd7408d498fd74fbd2623a92406b2fd3898faf0bf98f AS lint
 WORKDIR /src
-
 COPY go.mod go.sum ./
 RUN go mod download
-
+ARG CACHEBUST
 COPY . .
+RUN --mount=type=cache,target=/root/.cache/go-build \
+    --mount=type=cache,target=/root/.cache/golangci-lint \
+    golangci-lint run --config .golangci.yml ./...
 
-RUN make fmt-check
+# Test phase. -race needs cgo and so a C compiler, which the Debian Go
-RUN make lint
+# image ships and the alpine one does not. -count=1 because the mounted
+# cache also holds test results, which go test would otherwise replay.
+# -short skips the live RIS Live feed test, so the suite needs no network.
+# golang:1.27.2-bookworm, 2026-10-19
+FROM golang@sha256:5cf287a799e6b94384bad13d16b14904c531f51ba65792237e122ce42b392f61 AS test
+WORKDIR /src
+COPY go.mod go.sum ./
+RUN go mod download
+ARG CACHEBUST
+COPY . .
+RUN --mount=type=cache,target=/root/.cache/go-build \
+    go test -short -count=1 -timeout 90s -race -cover ./... || \
+    { echo "--- Rerunning with -v for details ---"; \
+      go test -short -count=1 -timeout 90s -race -v ./...; exit 1; }
 
-# Build stage
+# Build stage. Nothing is wanted from either phase above; the copies
-# golang:1.24-bookworm, 2026-09-21
+# are what make BuildKit build them first, so nothing after them can
-FROM golang@sha256:1a6d4452c65dea36aac2e2d606b01b4a029ec90cc1ae53890540ce6173ea77ac AS builder
+# run unless lint and test passed.
+# golang:1.27.2-bookworm, 2026-10-19
+FROM golang@sha256:5cf287a799e6b94384bad13d16b14904c531f51ba65792237e122ce42b392f61 AS builder
 
 # Install build dependencies (zstd for archive, gcc for CGO/sqlite3)
 RUN apt-get update && apt-get install -y --no-install-recommends \
@@ -26,36 +50,38 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
     && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /src
-
+COPY go.mod go.sum ./
-# Force BuildKit to run the lint stage before compiling or testing.
+RUN go mod download
+ARG CACHEBUST
 COPY --from=lint /src/go.sum /dev/null
-
+COPY --from=test /src/go.sum /dev/null
-# Copy everything
+# A tar-stream context keeps the sender's file owners, which git refuses.
+RUN git config --system --add safe.directory /src
 COPY . .
 
-# Vendor dependencies (must be after copying source)
+# Vendor dependencies for the source archive (must be after copying source)
-RUN go mod download && go mod vendor
+RUN go mod vendor
-
-# Run the test suite in the build stage: -race needs cgo and the C compiler
-# installed above. The suite is offline (the live-feed test is opt-in).
-RUN make test
 
 # Build the binary with CGO enabled (required for sqlite3). The version the
-# page footer shows is the VERSION build argument when one is given, otherwise
+# page footer shows is the VERSION build arg when one is given, otherwise
-# `git describe --tags --always` of the .git in the build context (git comes
+# `git describe --tags --always` on the .git in the build context: the tag
-# with this image): the tag on a tagged commit, tag-N-gHASH after one, the
+# on a tagged commit, tag-N-gHASH after one, the short commit when no tag is
-# short commit when no tag is reachable. A context that carries .git and still
+# reachable. With .git present, a version that is still empty, dev or
-# yields no version fails the build. The footer links to the full commit.
+# unknown fails the build: git is missing or could not read the checkout.
+# Without .git the footer says unknown. The footer links to the full commit.
 ARG VERSION
-RUN version="${VERSION:-$(git describe --tags --always || echo unknown)}"; \
+RUN --mount=type=cache,target=/root/.cache/go-build \
-    if [ -e .git ] && { [ -z "$version" ] || [ "$version" = dev ] || \
+    VERSION="${VERSION:-$(git describe --tags --always)}"; \
-        [ "$version" = unknown ]; }; then \
+    if [ -e .git ]; then \
-        echo "no version could be derived although the build context carries .git" >&2; \
+        case "$VERSION" in ""|dev|unknown) \
-        exit 1; \
+            echo "version is '$VERSION' although .git is present" >&2; \
+            exit 1 ;; \
+        esac; \
     fi; \
+    REVISION="$(git rev-parse --verify HEAD || echo unknown)"; \
     CGO_ENABLED=1 GOOS=linux go build -o /routewatch -ldflags "\
-        -X git.eeqj.de/sneak/routewatch/internal/version.GitRevision=$(git rev-parse --verify HEAD || echo unknown) \
+        -X git.eeqj.de/sneak/routewatch/internal/version.GitRevision=$REVISION \
-        -X git.eeqj.de/sneak/routewatch/internal/version.GitRevisionShort=$version" \
+        -X git.eeqj.de/sneak/routewatch/internal/version.GitRevisionShort=${VERSION:-unknown}" \
         ./cmd/routewatch
 
 # Create source archive with vendored dependencies
@@ -64,7 +90,7 @@ RUN tar --zstd -cf /routewatch-source.tar.zst \
     --exclude='*.tar.zst' \
     .
 
-# Runtime stage
+# Runtime stage, and the last one
 # debian:bookworm-slim, 2026-09-21
 FROM debian@sha256:3783cc01769c7b2b1b83a5c5ad96c815348e28ed7da68e2e3687004faa906251
 
@@ -76,6 +102,8 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
     curl \
     && rm -rf /var/lib/apt/lists/*
 
+ARG CACHEBUST
+
 # Create non-root user
 RUN useradd -r -u 1000 -m routewatch
 
@@ -111,8 +139,13 @@ EXPOSE 8080
 
 COPY ./entrypoint.sh /entrypoint.sh
 
-# Health check using the health endpoint, on the port PORT names
+# Health check using the health endpoint, on the port PORT names. During the
-HEALTHCHECK --interval=30s --timeout=5s --start-period=10s --retries=3 \
+# start period, Docker 25 and later checks every 5 seconds, and the first pass
+# makes the container healthy. After it, Docker checks every 30 seconds. The
+# start period covers the 60 seconds after which upaas reads the health
+# (README "Running under upaas"). Startup reads the routing table first,
+# which takes about 25 seconds when the table is full.
+HEALTHCHECK --interval=30s --timeout=5s --start-period=60s --retries=3 \
     CMD curl -sf "http://localhost:${PORT:-8080}/.well-known/healthcheck.json" || exit 1
 
 ENTRYPOINT ["/bin/bash", "/entrypoint.sh" ]

Makefile

@@ -52,7 +52,6 @@ clean:
 run: build
 	DEBUG=routewatch ./bin/routewatch 2>&1 | tee log.txt
 
+# Rebuild the AS list built into RouteWatch from as.csv (needs the network).
 asupdate:
-	@echo "Updating AS info data..."
+	@script/asupdate
-	@go run cmd/asinfo-gen/main.go | gzip > pkg/asinfo/asdata.json.gz.tmp && \
-		mv pkg/asinfo/asdata.json.gz.tmp pkg/asinfo/asdata.json.gz

README.md

@@ -1,10 +1,27 @@
 # RouteWatch
 
 RouteWatch is an MIT-licensed Go daemon by @sneak that monitors the BGP routing
-table in real time: it streams BGP UPDATE messages from the RIPE RIS Live
+table in real time. It works like this:
-service, maintains a live routing table in SQLite, and provides HTTP APIs for
-querying routing information.
 
+- It streams BGP UPDATE messages from the RIPE RIS Live service.
+- It keeps a live routing table in SQLite.
+- It serves web pages and a JSON API for looking routes up.
+
+## Rationale
+
+RIPE RIS Live streams the BGP updates that RIPE's route collectors receive, as
+they happen, from the moment a client connects. It reports changes, not the
+routing table they add up to, so it cannot say which AS an IP address is routed
+to now or which routes a prefix has. RouteWatch applies the feed's updates to
+live routing tables in its own SQLite database and answers IP-to-AS lookups and
+prefix pages from them.
+
+The feed sends no table when a client connects, so these tables hold only the
+routes the feed has announced since RouteWatch first started: an IP-to-AS lookup
+finds nothing for an address whose routes have not been announced since then.
+The tables also miss the updates RouteWatch drops when it falls behind under
+load and those sent while it is stopped, so a lookup or a prefix page can show a
+route the feed has withdrawn.
 
 ## Features
 
@@ -12,6 +29,8 @@ querying routing information.
 - Maintains live IPv4 and IPv6 routing tables
 - Tracks AS peering relationships
 - HTTP API for IP-to-AS lookups, prefix details, and AS information
+- IP lookups report the public blocklists an address is on, from copies kept in
+  memory and on disk
 - Automatic reconnection with exponential backoff
 - Batched database writes for high performance
 - Backpressure handling to prevent memory exhaustion
@@ -37,21 +56,171 @@ DEBUG=routewatch ./routewatch
 
 ## HTTP Endpoints
 
+Every route answers `HEAD` as it answers `GET`, without the body.
+
 ### Web Interface
-- `GET /` - Redirects to /status
+
-- `GET /status` - HTML status dashboard
+- `GET /` - Home page (HTML): headline figures, lookups by AS number and by IP
-- `GET /status.json` - JSON statistics
+  address, and a search of AS names. A card shows the `curl` command that looks
-- `GET /as/{asn}` - AS detail page (HTML)
+  up the visitor's own address at `/ip/{addr}`, and that command's live answer.
-- `GET /prefix/{prefix}` - Prefix detail page (HTML)
+- `GET /status` - HTML status dashboard, which reads `/api/v1/stats`
-- `GET /prefixlength/{length}` - IPv4 prefixes by mask length
+- `GET /status.json` - JSON statistics: feed, memory, database and route figures
-- `GET /prefixlength6/{length}` - IPv6 prefixes by mask length
+- `GET /.well-known/healthcheck.json` - Health check (JSON): 200 when the
-- `GET /ip/{ip}` - Redirects to prefix containing the IP
+  database answers and the RIS Live feed is connected, otherwise 503;
+  `data.status` is `ok` or `error` and `data.checks` says which check failed
+- `GET /as/{asn}` - AS detail page (HTML): the prefixes the AS originates and
+  its peers. When the database has no record of the AS, its handle and
+  description come from the AS list built into RouteWatch. The page then shows
+  no first-seen date.
+- `GET /prefix/{prefix}/{length}` - Prefix detail page (HTML), such as
+  `/prefix/192.0.2.0/24`: every live route for the prefix, one per peer, and its
+  origin ASes
+- `GET /prefixlength/{length}` - Up to 500 IPv4 prefixes with that mask length,
+  picked at random, each with its newest route (HTML)
+- `GET /prefixlength6/{length}` - The same for IPv6
+- `GET /ip/{addr}` - The same JSON answer as `/api/v1/ip/{ip}`
+- `GET /ip/` - The same, for the caller's own address: the address the request
+  came from or, when that is a trusted proxy (`TRUSTED_PROXIES`, see
+  Configuration), the client address in its `X-Forwarded-For`
 
 ### API v1
-- `GET /api/v1/stats` - Detailed statistics with handler metrics
+
-- `GET /api/v1/ip/{ip}` - Look up AS information for an IP address
+- `GET /api/v1/stats` - The `/status.json` figures plus connection, garbage
-- `GET /api/v1/as/{asn}` - Get prefixes announced by an AS
+  collector, stream and handler queue statistics. `stream.route_churn_per_min`
-- `GET /api/v1/prefix/{prefix}` - Get routes for a specific prefix
+  is the number of prefix announcements and withdrawals, from all peers, in the
+  last full clock minute.
+- `GET /api/v1/ip/{ip}` - Looks up each address. `{ip}` is an IP address, a host
+  name, or a comma-separated list of at most 10 of these. `results` has one
+  entry for each address, with:
+    - `ip` and `ptr`: the address and its PTR names
+    - `netblock` and the fields beside it: the address's most specific live
+      route, and that route's origin AS with its WHOIS details
+    - `blocklists`: the public blocklists the address is on (see Blocklists)
+
+    An address with no live route has a result without the route and AS fields.
+    Of the addresses a host name resolves to, only public ones are looked up. A
+    name with none, such as one that resolves to a private address, is reported
+    as `host name has no public address`, without its addresses. `errors` has
+    one entry for each host name that failed. The answer is 400 when:
+    - `{ip}` lists more than 10
+    - `{ip}` comes to more than 10 addresses once its names are resolved
+    - nothing in `{ip}` is an IP address or a host name with a public address
+
+    A PTR lookup gets 2 seconds. An address whose lookup takes longer is
+    answered without PTR names.
+
+- `GET /api/v1/as/{asn}` - The AS and the prefixes it originates, IPv4 and IPv6
+  apart, each with when its newest route was updated. When the database has no
+  record of the AS, its handle and description come from the AS list built into
+  RouteWatch. The answer then has no `first_seen` or `last_seen`.
+- `GET /api/v1/prefix/{prefix}/{length}` - Every live route for the prefix, one
+  per peer, and the same routes grouped by origin AS in `origins`
+- `GET /api/v1/search/as?q={text}` - Finds the ASes whose handle or description
+  contains `{text}`, ignoring case. `data` holds:
+    - `query`: `{text}` without its leading and trailing spaces
+    - `total`: the number of ASes that match
+    - `results`: the first 50 of them in AS number order, each with `asn`,
+      `handle` and `description`
+
+    The search reads the AS list built into RouteWatch, not the database. The
+    list comes from
+    [ipverse/as-metadata](https://github.com/ipverse/as-metadata). A match can
+    therefore be an AS that the database has no record of. A `q` that is
+    missing, empty or only spaces answers 400.
+
+A JSON answer is `{"status": "ok", "data": ...}`, or for an error
+`{"status": "error", "error": {"msg": ..., "code": ...}}`, and carries an
+`@meta` object with the time zone, the API version and the time the request
+took. The health check is the exception: its 503 answer is also
+`{"status": "ok", "data": ...}`, with `error` in `data.status`.
+
+A request that runs past its time limit answers 504: the limit is 30 seconds,
+and 4 seconds for the database reads behind `/status.json` and `/api/v1/stats`.
+An AS number, prefix or mask length that does not parse answers 400, as does a
+mask length out of range for its IP version. An AS answers 404 when it is not in
+the AS list built into RouteWatch and the database has neither a record of it
+nor a live route it originates. AS 0 always answers 404, because it means no
+known origin. The AS name search never lists it. A prefix with no live route
+answers 404 too. The HTML pages give these errors as plain text. Under `/api/v1`
+and `/ip/`, a path no route matches answers 404 with the message `Not found`,
+and `POST`, `PUT`, `PATCH`, `DELETE`, `OPTIONS`, `TRACE` or `CONNECT` answers
+405 with `Method not allowed`. A 500 answer says only `Internal server error`,
+and a host name that does not resolve only `could not resolve host name`; the
+detail goes to the log.
+
+Every answer carries `Strict-Transport-Security`, `Content-Security-Policy`,
+`X-Frame-Options`, `X-Content-Type-Options`, `Referrer-Policy` and
+`Permissions-Policy`. The Content-Security-Policy lets a page load content only
+from the server, and does not allow inline content with `'unsafe-inline'`: the
+pages keep their CSS and JavaScript in inline `<style>` and `<script>` blocks,
+so each answer's policy names a random nonce made for that answer, and the
+browser applies or runs only the blocks that carry it. The pages therefore use
+no `style` attributes, which a nonce cannot cover.
+
+An origin AS of 0 (`origin_asn` in a route, `asn` in an IP lookup, the key `0`
+in a prefix's `origins`) means the route has no known origin: its AS path ends
+in an AS set of several members, any of which may have originated it. The pages
+show such a route as having no known origin.
+
+## Blocklists
+
+The IP lookup reports which public blocklists list each address. Each result has
+a `blocklists` list with one entry for each list the address is on:
+
+- `name`: the list's name, from the lists below
+- `description`: what being on the list means
+- `updated`: when RouteWatch last downloaded the list, or last heard from its
+  publisher that its copy was current
+
+An address on no list has an empty `blocklists`. An address with no live route
+still has a result with its `blocklists`.
+
+The lists:
+
+- `dshield`: the 20 networks that DShield saw scan the most targets in the last
+  three days. From [DShield](https://isc.sans.edu/feeds_doc.html), downloaded
+  every hour.
+- `feodo_tracker`: botnet command-and-control servers. From abuse.ch
+  [Feodo Tracker](https://feodotracker.abuse.ch/blocklist/), downloaded every 15
+  minutes.
+- `firehol_abusers_30d`: addresses reported for comment, form or forum spam in
+  the last 30 days. From
+  [FireHOL](https://iplists.firehol.org/?ipset=firehol_abusers_30d), downloaded
+  every 5 hours.
+- `firehol_level3`: addresses reported for attacks, spyware or viruses in the
+  last 30 days. From
+  [FireHOL](https://iplists.firehol.org/?ipset=firehol_level3), downloaded every
+  3 hours.
+- `spamhaus_drop` and `spamhaus_dropv6`: IPv4 and IPv6 networks that Spamhaus
+  says spammers or cyber criminals have hijacked or lease. From
+  [Spamhaus DROP](https://www.spamhaus.org/blocklists/do-not-route-or-peer/),
+  downloaded once a day. Spamhaus merged its EDROP list into DROP.
+- `tor_exits`: Tor exit nodes, where traffic from the Tor network leaves. From
+  the [Tor Project](https://check.torproject.org/torbulkexitlist), downloaded
+  every hour.
+
+No list is downloaded more often than its publisher updates it or allows. Each
+can be downloaded whole without an account or a key.
+
+How RouteWatch keeps the lists:
+
+- It holds each list in memory, so a lookup makes no outside request.
+- It keeps a copy of each list in the `blocklists` directory of the state
+  directory (see Configuration). Each copy is a file named after its list.
+- Beside each copy, a file ending in `.validators` keeps the `Last-Modified` and
+  `ETag` the publisher sent with that copy.
+- At startup it loads each list from its copy. It downloads a list at once when
+  the copy is older than the list's interval, or when there is no copy.
+- A download sends the copy's `Last-Modified` as `If-Modified-Since` and its
+  `ETag` as `If-None-Match`. The publisher then sends the list only when it has
+  changed. Without either, the download asks for the whole list.
+- A download fails when it takes over a minute or its file is over 16 MiB.
+- When a download fails, or its file does not parse, RouteWatch keeps the
+  previous copy and logs the failure. It tries again one interval later.
+
+A successful IP lookup answer carries `Cache-Control` with a `max-age` of 15
+minutes, the shortest list interval. The answer for the caller's own address
+(`/ip/`) is `private`, and the others are `public`.
 
 ## Code Structure
 
@@ -62,15 +231,19 @@ routewatch/
 │   ├── asinfo-gen/         # Utility to generate AS info data
 │   └── streamdumper/       # Debug utility for raw stream output
 ├── internal/
+│   ├── blocklist/          # Public IP blocklists: downloads and lookups
 │   ├── routewatch/         # Core application logic
 │   ├── server/             # HTTP server and handlers
 │   ├── database/           # SQLite storage layer
+│   ├── routetable/         # The whole routing table in memory
 │   ├── streamer/           # RIPE RIS Live client
 │   ├── ristypes/           # BGP message data structures
 │   ├── logger/             # Structured logging wrapper
 │   ├── metrics/            # Performance metrics tracking
 │   ├── config/             # Configuration management
-│   └── templates/          # HTML templates
+│   ├── templates/          # HTML templates
+│   ├── version/            # Build version information
+│   └── whois/              # WHOIS lookups of AS details
 └── pkg/
     └── asinfo/             # AS information lookup (public API)
 ```
@@ -89,7 +262,7 @@ routewatch/
 │  │   Streamer   │───▶│   Handlers   │───▶│   Database   │       │
 │  │              │    │              │    │              │       │
 │  │ RIS Live     │    │ - ASHandler  │    │ SQLite with  │       │
-│  │ WebSocket    │    │ - PeerHandler│    │ WAL mode     │       │
+│  │ HTTP stream  │    │ - PeerHandler│    │ WAL mode     │       │
 │  │ client       │    │ - PrefixHdlr │    │              │       │
 │  │              │    │ - PeeringHdlr│    │ Tables:      │       │
 │  └──────────────┘    └──────────────┘    │ - asns       │       │
@@ -111,24 +284,39 @@ routewatch/
     - Signal handlers registered for graceful shutdown
 
 2. **Initialization** (`internal/routewatch/app.go`)
-   - Database created with SQLite schema (WAL mode, 3GB cache)
+    - Database created with SQLite schema (WAL mode, page caches as in "Memory")
+    - The routing table read from the database into memory, before the streamer
+      connects
     - Message handlers registered with the streamer
     - HTTP server started on configured port
 
 3. **Message Processing Pipeline**
+
     ```
     RIS Live Stream → JSON Parser → Message Dispatcher → Handler Queues → Batch Writers → SQLite
     ```
+
     - Streamer connects to `ris-live.ripe.net` via HTTP
     - Parses BGP UPDATE messages from JSON stream
     - Dispatches to registered handlers based on message type
     - Each handler has its own queue with backpressure handling
-   - Handlers batch writes for efficiency (25K-30K ops, 1-2s timeout)
+    - Handlers batch writes for efficiency: the AS, peer and prefix handlers
+      write 30,000, 10,000 and 25,000 updates at a time, or after 2, 2 and 1
+      seconds; the peering handler writes the peerings it has collected every 30
+      seconds
 
 4. **Handler Details**
     - **ASHandler**: Tracks all ASNs seen in AS paths
     - **PeerHandler**: Records BGP peer information
-   - **PrefixHandler**: Maintains live routing table (upserts on announcement, deletes on withdrawal)
+    - **PrefixHandler**: Maintains live routing table (upserts on announcement,
+      deletes on withdrawal or on an announcement whose AS path is empty or
+      holds AS 0 anywhere, deletes all of a peer's routes when its BGP session
+      goes down). Routes are kept per peer IP, so when one peer IP has sessions
+      on two collectors, a session going down on one also removes the routes the
+      other still carries, until they are announced again. It also applies each
+      update to the routing table in memory, and writes the table's changes to
+      the database every second. A write takes at most 20,000 changed prefix and
+      origin AS pairs, and the rest follow in the next writes.
     - **PeeringHandler**: Extracts AS peering relationships from AS paths
 
 5. **HTTP Request Flow**
@@ -138,55 +326,91 @@ routewatch/
 
 ### Key Design Patterns
 
-- **Batched Writes**: All database operations are batched for performance
+- **Batched Writes**: The handlers write the feed's updates to the database in
-- **Backpressure**: Probabilistic message dropping when queues exceed 50% capacity
+  batches for performance
-- **Graceful Shutdown**: 60-second timeout, flushes all pending batches
+- **Backpressure**: Probabilistic message dropping when queues exceed 50%
-- **Reconnection**: Exponential backoff (5s-320s) with reset after 30s of stable connection
+  capacity
+- **Graceful Shutdown**: on SIGTERM or SIGINT the streamer stops, the handlers
+  write the batches they hold, the HTTP server stops and the database is
+  checkpointed and closed. The daemon gives this 8 seconds and exits with status
+  0; "Running under upaas" says what a stop loses, logs and leaves undone when
+  the 8 seconds run out.
+- **Reconnection**: Any end of the stream other than shutdown, including RIS
+  Live ending the response normally, marks the feed disconnected and reconnects
+  with exponential backoff (5s-320s). The backoff resets after a stream that
+  lasted more than 30s. A stream with no line for 60s is dropped and
+  reconnected.
 - **IP Lookup**: the most specific live route is found by looking up the
-  address's prefix at each mask length, longest first, in the prefix index
+  address's prefix at each mask length, longest first, in the prefix index (at
-  (at most 33 lookups for IPv4, 129 for IPv6); prefixes are stored in the
+  most 25 lookups for IPv4, 113 for IPv6); prefixes are stored in the text form
-  text form Go's `net/netip` prints
+  Go's `net/netip` prints. Routes shorter than /8 for IPv4 or /16 for IPv6, such
+  as a default route, are not used.
 
 ### Database Schema
 
 ```sql
 -- Core tables
-asns(id, number, handle, description, first_seen, last_seen)
+asns(asn PRIMARY KEY, handle, description,
-prefixes_v4(id, prefix, mask_length, first_seen, last_seen)
+     as_name, org_name, org_id, address, country_code, abuse_email,
-prefixes_v6(id, prefix, mask_length, first_seen, last_seen)
+     abuse_phone, tech_email, tech_phone, rir, rir_registration_date,
+     rir_last_modified, whois_raw,
+     first_seen, last_seen, whois_updated_at)
+prefixes_v4(prefix PRIMARY KEY, first_seen) WITHOUT ROWID
+prefixes_v6(prefix PRIMARY KEY, first_seen) WITHOUT ROWID
 
--- Live routing tables (one per IP version)
+-- Live routing tables (one per IP version), one route per peer and prefix
-live_routes_v4(id, prefix, mask_length, origin_asn, peer_ip, as_path,
+live_routes_v4(id INTEGER PRIMARY KEY, prefix, mask_length, origin_asn,
-               next_hop, last_updated)
+               peer_ip, as_path, next_hop, last_updated,
-live_routes_v6(id, prefix, mask_length, origin_asn, peer_ip, as_path,
+               UNIQUE(peer_ip, prefix))
-               next_hop, last_updated)
+live_routes_v6(id INTEGER PRIMARY KEY, prefix, mask_length, origin_asn,
+               peer_ip, as_path, next_hop, last_updated,
+               UNIQUE(peer_ip, prefix))
 
 -- Relationship tracking
-peerings(id, as_a, as_b, first_seen, last_seen)
+peerings(id PRIMARY KEY, as_a, as_b, first_seen, last_seen,
-bgp_peers(id, peer_ip, peer_asn, last_message_type, last_seen)
+         UNIQUE(as_a, as_b))
+bgp_peers(id PRIMARY KEY, peer_ip UNIQUE, peer_asn, first_seen, last_seen,
+          last_message_type)
+
+-- Nothing in the daemon writes this table
+announcements(id PRIMARY KEY, prefix_id, peer_asn, origin_asn, path,
+              next_hop, timestamp, is_withdrawal)
 ```
 
+The indexes are in `internal/database/schema.sql`.
+
 ## Configuration
 
 Configuration is handled via environment variables and OS-specific paths:
 
 | Variable           | Default                          | Description                                                 |
-|----------|----------|-------------|
+| ------------------ | -------------------------------- | ----------------------------------------------------------- |
 | `PORT`             | `8080`                           | HTTP server port, a whole number from 1 to 65535            |
 | `DEBUG`            | (empty)                          | Set to `routewatch` for debug logging                       |
 | `XDG_DATA_HOME`    | `/var/lib` (in the Docker image) | Base of the state directory; must be an absolute path       |
 | `GOMEMLIMIT`       | `1536MiB` (in the Docker image)  | Go soft memory limit; see Memory                            |
 | `MALLOC_ARENA_MAX` | `2` (in the Docker image)        | glibc malloc arena cap, a positive whole number; see Memory |
+| `TRUSTED_PROXIES`  | (empty)                          | Addresses and CIDR ranges of trusted proxies; see below     |
 
 A variable that is set to an invalid value stops the start with an error and a
 non-zero exit. An empty variable counts as unset.
 
-State directory (database location):
+`TRUSTED_PROXIES` lists the reverse proxies whose `X-Forwarded-For` the server
-- macOS: `~/Library/Application Support/routewatch/`
+believes, comma-separated, such as `10.0.0.0/8,192.0.2.1`. For a request from
-- Linux: `/var/lib/berlin.sneak.app.routewatch/` when running as root,
+one of them, the request log and `/ip/` use the last address in
-  otherwise `$XDG_DATA_HOME/berlin.sneak.app.routewatch/` (with `XDG_DATA_HOME`
+`X-Forwarded-For` that is not a trusted proxy: each proxy adds the address it
-  unset, `~/.local/share/berlin.sneak.app.routewatch/`). In the Docker image
+received the request from to the end, so the entries before it are whatever the
-  this is `/var/lib/berlin.sneak.app.routewatch/`.
+client sent. For a request from any other address they use the address the
+request came from, whatever forwarding headers it carries. `X-Real-IP` and
+`True-Client-IP` are not read.
+
+State directory (the database and the blocklist copies):
+
+- macOS: `~/Library/Application Support/berlin.sneak.app.routewatch/`
+- Linux: `/var/lib/berlin.sneak.app.routewatch/` when running as root, otherwise
+  `$XDG_DATA_HOME/berlin.sneak.app.routewatch/` (with `XDG_DATA_HOME` unset,
+  `~/.local/share/berlin.sneak.app.routewatch/`). In the Docker image this is
+  `/var/lib/berlin.sneak.app.routewatch/`.
 
 ## Memory
 
@@ -194,9 +418,14 @@ The daemon holds a live routing table, so its memory grows with the size of the
 data it tracks. The image sets ceilings that keep it inside a 5 GiB container.
 
 Budget:
+
 - Go heap: a 1.5 GiB soft limit (`GOMEMLIMIT=1536MiB`, set in the image).
-- SQLite: at most 640 MiB of page cache across the connection pool (64 MiB per
+- SQLite: 1 GiB of page cache for the one connection every write uses, which
-  connection, 10 connections) and a 1.5 GiB hard heap limit for the C library.
+  keeps the pages the route writes touch from one batch to the next, and 16 MiB
+  for each of the 9 reading connections, whose cache SQLite empties whenever
+  another connection has written: about 1.15 GiB in all. A 1.5 GiB soft heap
+  limit sits above the caches, so it does not shrink them, and a 2 GiB hard heap
+  limit caps the C library.
 - glibc allocator: the SQLite C library runs on glibc `malloc`, which frees
   page-cache chunks back to per-arena free lists rather than to the kernel, so
   process RSS tracks the high-water mark of those arenas, not SQLite's live
@@ -206,6 +435,9 @@ Budget:
   database writes are already serialized.
 - About 0.2 GiB for everything else in the runtime.
 
+The ceilings add up to 3.7 GiB: the Go heap's 1.5 GiB, SQLite's 2 GiB and the
+0.2 GiB for the rest.
+
 Run the container with a memory limit of 5 GiB and swap disabled:
 
 ```bash
@@ -216,19 +448,37 @@ or the equivalent in your deployment tool. This leaves headroom above the
 ceilings for spikes and the kernel page cache.
 
 Override the Go soft limit by setting `GOMEMLIMIT` in the environment (for
-example `-e GOMEMLIMIT=1GiB`); this replaces the image default. `MALLOC_ARENA_MAX`
+example `-e GOMEMLIMIT=1GiB`); this replaces the image default.
-can be overridden the same way, but raising it lets RSS climb again on a
+`MALLOC_ARENA_MAX` can be overridden the same way, but raising it lets RSS climb
-many-core host.
+again on a many-core host.
 
 What happens at each limit:
+
 - Go soft limit: as the heap approaches `GOMEMLIMIT`, the runtime runs garbage
   collection more aggressively rather than growing further.
-- SQLite: at a 1 GiB soft heap limit it recycles its page cache instead of
+- SQLite: at the 1.5 GiB soft heap limit it recycles its page cache instead of
-  allocating more; at the 1.5 GiB hard heap limit a statement fails with an
+  allocating more; at the 2 GiB hard heap limit a statement fails with an
   out-of-memory error, and the handler logs it and drops that batch. The process
   keeps running.
 - Handler queues: each of the four handler queues holds at most 20,000 messages.
-  When a queue fills, the streamer drops messages instead of blocking.
+  When a queue fills, the streamer drops messages instead of blocking, except a
+  `STATE` message, which it holds, in feed order, until there is room: a few
+  hundred bytes each, about ten a second.
+- Handler batches: while the AS, peer or prefix handler writes a batch to the
+  database, the updates that arrive collect in its next batch, which stops at
+  ten times the batch size: 300,000 updates for the AS handler
+  (`asnBatchMaxSize`, about 10 MiB), 100,000 for the peer handler
+  (`peerBatchMaxSize`, about 10 MiB) and 250,000 for the prefix handler
+  (`prefixBatchMaxSize`, about 60 MiB). At that size the handler takes nothing
+  more off its queue until the write ends, so the queue fills and the streamer
+  drops messages as above. When the delete of a peer's live routes after its
+  session went down fails, the prefix handler keeps the updates the peer sent
+  after the down for the next write, which retries the delete first; on top of
+  its 250,000 it keeps at most the last 25,000 of them (`prefixBatchSize`, about
+  6 MiB) and drops and logs the older ones, so while the delete keeps failing
+  the other peers' updates still come off the queue. With the batches being
+  written, the three handlers' batches take at most about 0.25 GiB of the Go
+  heap's 1.5 GiB.
 
 With `DEBUG=routewatch` the daemon logs a `System stats` line every 60 seconds
 with the goroutine count and Go memory figures.
@@ -239,15 +489,38 @@ What the [upaas](https://git.eeqj.de/sneak/upaas) app needs:
 
 - Container port: `8080`.
 - Volume: one, at container path `/var/lib/berlin.sneak.app.routewatch`.
-- Environment: nothing is required. Leave `XDG_DATA_HOME`, `GOMEMLIMIT` and
+- Environment: `TRUSTED_PROXIES`, set to the address the reverse proxy in front
-  `MALLOC_ARENA_MAX` at the image's values. `DEBUG=routewatch` is optional and
+  of the app connects from, as the container sees it; without it the request log
-  adds the `System stats` memory line to the log.
+  and `/ip/` give the proxy's address for every request. Nothing else is needed:
+  leave `XDG_DATA_HOME`, `GOMEMLIMIT` and `MALLOC_ARENA_MAX` at the image's
+  values. `DEBUG=routewatch` is optional and adds the `System stats` memory line
+  to the log.
 - Memory Limit: `5g`, the 5 GiB limit from Memory above. upaas sets no swap
   limit, so on a host with swap Docker allows the same amount of swap again.
 - Health check: the image's `HEALTHCHECK` requests
   `/.well-known/healthcheck.json` on the container port. upaas reads the
-  container's health 60 seconds after a deploy and fails the deploy unless it
+  container's health 60 seconds after a deploy and fails the deploy unless it is
-  is `healthy`.
+  `healthy`. For the first 60 seconds, Docker 25 and later runs the check every
+  5 seconds, and the first pass makes the container healthy. Startup first reads
+  the routing table, which takes about 25 seconds when the table is full.
+- Stop grace: the container needs 9 seconds between SIGTERM and SIGKILL. upaas
+  gives 10 (`stopTimeoutSeconds` in its `internal/docker/client.go`), as
+  `docker stop` does by default. The daemon allows 8 seconds for the handlers to
+  write what they hold and for the requests in progress to finish, then stops
+  the HTTP server, closes the database and exits with status 0. A longer grace
+  changes nothing. What is not written after 8 seconds is lost, and the streamer
+  and each handler log how much. When the prefix handler has fallen behind the
+  feed, a stop loses far more: its batch is full at 250,000 updates and the
+  streamer's worker for it waits for room, so the streamer's stop takes the
+  whole 8 seconds and every handler then skips its last write. Lost then are the
+  messages left on the four handlers' queues, what the AS, peer and peering
+  handlers hold, the prefix handler's batch, and the write it has in progress,
+  up to another 250,000 updates, which take about 20 seconds and are cut off by
+  the exit; withdrawals and peer downs are among them. A route withdrawn in them
+  stays in the live route table until its peer announces or withdraws the prefix
+  again or its session goes down. When the 8 seconds run out the database is not
+  closed, and SQLite rolls back a write the exit cut off when the database next
+  opens.
 
 ## Development
 
@@ -262,7 +535,7 @@ make fmt
 make lint
 
 # Build
-make
+make build
 ```
 
 ## Entrypoints
@@ -270,35 +543,45 @@ make
 This repository adheres to the
 [Scripts to Rule Them All](https://github.com/github/scripts-to-rule-them-all)
 standard: normalized scripts in `script/` are the entrypoints for the
-development workflow, and the Makefile targets are thin shims that call
+development workflow, and the Makefile targets are thin shims that call them. We
-them. We provide:
+provide:
 
-- `script/bootstrap` — install all development dependencies (go,
+- `script/bootstrap` — install all development dependencies (git, make, go, the
-  golangci-lint, Go module download)
+  pinned node and yarn, prettier from `yarn.lock`, Go module download)
 - `script/setup` — make a fresh clone ready for development: runs
   `script/bootstrap`, then `script/install-precommit`
-- `script/projectname` — print the project name (used for the Docker
+- `script/projectname` — print the project name (used for the Docker image tag)
-  image tag)
+- `script/test` — build the `test` phase of the `Dockerfile` alone, which runs
-- `script/test` — run the test suite
+  `go test -short -count=1 -timeout 90s -race -cover ./...` (verbose rerun on
-  (`go test -short -timeout 30s -race -cover ./...`, verbose rerun on
+  failure). The `-short` flag skips the live-network integration test so the run
-  failure). The `-short` flag skips the live-network integration test so the
+  is deterministic and offline.
-  default run is deterministic and offline.
+- `script/lint` — build the `lint` phase of the `Dockerfile` alone, which runs
-- `script/lint` — run `go vet ./...` and `golangci-lint run`
+  golangci-lint v2.14.0 with `.golangci.yml`
-- `script/fmt` — format all code (writes)
+- `script/fmt` — format the Go code with `go fmt` and the markdown with prettier
+  (writes)
 - `script/fmt-check` — check formatting (read-only)
-- `script/check` — run `script/test`, `script/lint`, and
+- `script/check` — run `script/test`, `script/lint`, and `script/fmt-check`
-  `script/fmt-check`
+- `script/docker` — build the Docker image tagged via `script/projectname`
-- `script/docker` — build the Docker image tagged via
+- `script/cibuild` — CI entrypoint: runs `script/bootstrap` and `script/check`,
-  `script/projectname`
+  then builds the image
-- `script/cibuild` — CI entrypoint: `docker build .`
+- `script/precommit` — pre-commit gate: `go mod tidy` (must not change `go.mod`
-- `script/precommit` — pre-commit gate: `go mod tidy` + `go fmt` (must
+  or `go.sum`) and `go fmt` (writes), then `script/check`
-  not change files), then `script/check`
+- `script/install-precommit` — install the git pre-commit hook that runs
-- `script/install-precommit` — install the git pre-commit hook that
+  `script/precommit`
-  runs `script/precommit`
+- `script/asupdate` — rebuild the AS list built into RouteWatch,
+  `pkg/asinfo/asdata.json.gz`, from the `as.csv` file of
+  [ipverse/as-metadata](https://github.com/ipverse/as-metadata). It needs the
+  network. It runs `cmd/asinfo-gen`, which fails when it finds no AS. When
+  `cmd/asinfo-gen` fails, the old list stays and the script exits non-zero.
 
-The live-network integration test `TestRouteWatchLiveFeed` streams the RIPE
+Tests and the linter run only in Docker, as the `test` and `lint` phases of the
-RIS feed for a few seconds and is skipped in short mode. To run it on demand,
+`Dockerfile`; the image cannot be built unless both pass. Every `docker build`
-invoke `go test` directly without `-short`:
+in `script/` passes a new `CACHEBUST` value, so each run executes every check,
+while Go's build cache and golangci-lint's cache are kept between builds.
+
+The live-network integration test `TestRouteWatchLiveFeed` streams the RIPE RIS
+feed for a few seconds and is skipped in short mode. To run it on demand, invoke
+`go test` directly without `-short`:
 `go test -run TestRouteWatchLiveFeed ./internal/routewatch/`.
 
 ## License

REPO_POLICIES.md

@@ -1,6 +1,6 @@
 ---
 title: Repository Policies
-last_modified: 2026-07-06
+last_modified: 2026-10-18
 ---
 
 This document covers repository structure, tooling, and workflow standards. Code
@@ -54,21 +54,37 @@ style conventions are in separate documents:
   `cibuild`. `script/bootstrap` installs all dependencies idempotently and
   assumes nothing is present: base tools come from nix, apt, brew, or apk
   (detected in that order; apt runs noninteractive). For node it uses the
-  installed node if present; otherwise it installs a PINNED node version via
+  installed node only when its major version is the pinned one; otherwise (node
-  nvm, first installing nvm itself if missing — from a hash-verified GitHub
+  missing, or another major, such as the node 24 the CI runner image ships) it
-  release archive (never `curl | sh`), with bash installed as an explicit
+  installs the PINNED node version via nvm and installs yarn and the packages
-  prerequisite since nvm requires bash. yarn is then pinned via
+  under it. nvm itself is installed first if missing, from a hash-verified
-  `corepack prepare yarn@<version> --activate`. Never install "latest" or "lts";
+  GitHub release archive (never `curl | sh`), with bash installed as an explicit
-  always exact versions. `script/cibuild` runs the CI build: it changes to the
+  prerequisite since nvm requires bash. Only the major version is compared
-  repo root and runs `docker build .`; the Gitea workflow calls it. Four further
+  because the `Dockerfile` stages start from a node image whose exact version is
-  scripts are our own extensions to the standard: `script/check` runs
+  not the pin, and nvm cannot install a prebuilt node on alpine. yarn is pinned
-  `script/test`, `script/lint`, and `script/fmt-check`; `script/precommit` is
+  via `corepack prepare yarn@<version> --activate`. Never install "latest" or
-  what the git pre-commit hook runs, and it calls `script/check`;
+  "lts"; always exact versions. `script/cibuild` runs the CI build: it changes
-  `script/install-precommit` installs the git pre-commit hook (the `make hooks`
+  to the repo root, runs `script/bootstrap`, runs `script/check`, and builds the
-  target shims to it); and `script/projectname` (literally that filename) simply
+  image with the version; the Gitea workflow calls it. **`script/cibuild` runs
-  outputs the project's name. Scripts that need the name call
+  `script/bootstrap` first**, because the workflow checks out the repo and runs
-  `script/projectname` — e.g. `script/docker` assembles its image tag from it —
+  nothing else, while `script/fmt-check` runs the formatter on the host: on a
-  so those scripts stay byte-identical across all repos. Repo-type-specific
+  pristine checkout with nothing installed the run dies there, after the
+  containerised gates have passed. **The bootstrap alone is not enough**: when
+  `script/bootstrap` installs node and yarn under nvm it leaves neither on the
+  `PATH` of the shell that called it, so a bare `yarn` either exits 127 or runs
+  under another node. The host entrypoints that need yarn — `script/fmt` and
+  `script/fmt-check` — therefore use the same test as `script/bootstrap`: when
+  the node on `PATH` has the pinned major version they run the `yarn` on `PATH`,
+  and otherwise they load nvm through `$HOME/.nvm/nvm.sh` and run yarn under the
+  pinned node. A runner carrying nothing but docker and git then gets through
+  `script/check`. Four further scripts are our own extensions to the standard:
+  `script/check` runs `script/test`, `script/lint` and `script/fmt-check`;
+  `script/precommit` is what the git pre-commit hook runs, and it calls
+  `script/check`; `script/install-precommit` installs the git pre-commit hook
+  (the `make hooks` target shims to it); and `script/projectname` (literally
+  that filename) simply outputs the project's name. Scripts that need the name
+  call `script/projectname` — e.g. `script/docker` assembles its image tag from
+  it — so those scripts stay byte-identical across all repos. Repo-type-specific
   pre-commit extras (e.g. `go mod tidy` verification in Go repos) belong in
   `script/precommit`, not in the hook itself. Model scripts are at
   `https://git.eeqj.de/sneak/prompts/raw/branch/main/script/<name>`. The README
@@ -89,87 +105,290 @@ style conventions are in separate documents:
   contributor should be able to understand the entire development workflow by
   reading the Makefile.
 
-- Every repo should have a `Dockerfile`. All Dockerfiles must run `make check`
+- Every repo should have a `Dockerfile`, and it carries the repo's gates: a
-  as a build step so the build fails if the branch is not green. For non-server
+  `lint` phase and a `test` phase, with the final stage depending on both so the
-  repos, the Dockerfile should bring up a development environment and run
+  image cannot be built unless they pass. For non-server repos the final stage
-  `make check`. For server repos, `make check` should run as an early build
+  brings up a development environment; for server repos it is the runtime image.
-  stage before the final image is assembled. Dockerfiles install development
+  The gate phases and the build stage start from their pinned base images and
-  prerequisites by running `script/bootstrap` rather than duplicating installs
+  install what those images lack either inline, as the canonical Go `Dockerfile`
-  inline; COPY `script/` and the dependency manifests (`package.json` +
+  below does for `git`, or by running `script/bootstrap`, as the `prompts`
-  `yarn.lock`, `go.mod` + `go.sum`, etc.) before running it so the bootstrap
+  repo's own `Dockerfile` does for its yarn packages. The development
-  layer stays cached until dependencies change.
+  environment stage installs development prerequisites by running
+  `script/bootstrap` rather than duplicating its installs inline. A stage that
+  runs `script/bootstrap` COPYs `script/` and the dependency manifests
+  (`package.json` + `yarn.lock`, `go.mod` + `go.sum`, etc.) before running it.
 
-- **Dockerfiles must use a separate lint stage for fail-fast feedback.** Go
+- **Linting and testing run in Docker, as phases of the `Dockerfile`.** There is
-  repos use a multistage build where linting runs in an independent stage based
+  no separate lint file. `script/lint` and `script/test` each build one phase
-  on the `golangci/golangci-lint` image (pinned by hash). This stage runs
+  and nothing else:
-  `make fmt-check` and `make lint` before the full build begins. The build stage
-  then declares an explicit dependency on the lint stage via
-  `COPY --from=lint /src/go.sum /dev/null`, which forces BuildKit to complete
-  linting before proceeding to compilation and tests. This ensures lint failures
-  surface in seconds rather than minutes, without blocking on dependency
-  download or compilation in the build stage.
 
-    The standard pattern for a Go repo Dockerfile is:
+    ```sh
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
+        --target lint --output type=cacheonly .
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
+        --target test --output type=cacheonly .
+    ```
+
+    **A stage that is not the last one in the file is built only when the final
+    stage's chain depends on it, or when `--target` names it.** That is why the
+    two gates are always invoked by name here, and why the final stage carries a
+    `COPY --from=` of a harmless file from each of them: without that edge a
+    plain `docker build .` builds the last stage alone and exits 0 having linted
+    and tested nothing.
+
+    **The gate builds write no image.** With `--output type=cacheonly` the phase
+    runs and a failing step fails the build, but the result is not exported.
+    Nothing uses those images, and writing one out is slow: a Go test phase's
+    image holds the toolchain. A build given neither `--output` nor `-t` writes
+    an untagged image and leaves it dangling, on every developer host and every
+    CI runner. `script/cibuild` and `script/docker` build the image that ships
+    and tag it, so each build replaces the previous image; each assigns the tag
+    on its own line before the build, so `set -e` stops it where
+    `script/projectname` fails.
+
+    Inside a phase the tool is invoked directly — `golangci-lint`, `go test`,
+    `eslint`, `prettier` — never through `make lint` or `script/test`, which are
+    themselves a `docker build` and would recurse into a daemon that does not
+    exist in a build step. Formatting is the exception and stays on the host:
+    `script/fmt` writes the working tree, and `script/fmt-check` is its
+    read-only twin.
+
+    **No lint verdict may come from a host invocation of the linter.** On a
+    shared host golangci-lint reads a result cache keyed on file content rather
+    than location, so a second checkout of the same content is served the first
+    one's findings, and a host-global lock in `$TMPDIR` makes concurrent runs
+    exit non-zero with `parallel golangci-lint is running` — a status a caller
+    cannot tell from real findings. Both have produced wrong verdicts in this
+    org, in both directions. A container has its own `TMPDIR`, so the lock is
+    out of reach. The lint phase does keep golangci-lint's cache between check
+    builds (below), but it lints every checkout at `/src`, and a finding is
+    taken from that cache only for the same files, settings and golangci-lint
+    version.
+
+- **Any build that runs checks is given a new `CACHEBUST` value.** Docker
+  invalidates a `COPY` layer only when the copied content changes, so on an
+  unchanged tree the check `RUN` is served from cache, nothing executes, and the
+  build still exits 0. Every `docker build` in `script/` therefore passes
+  `--build-arg CACHEBUST="$(date +%s)-$$"`, the time and the script's process
+  id, so no two builds share a value, not even the three in one
+  `script/cibuild`. `script/lint`, `script/test`, `script/cibuild` and
+  `script/docker` are the four, and there is no fifth — `script/check` runs the
+  two gate phases and `script/fmt-check`, and builds no image of its own. In
+  every stage of the `Dockerfile`, `ARG CACHEBUST` comes right after the steps
+  that install system packages and download Go's modules, or right after `FROM`
+  in a stage with neither, and every step that lints, tests or compiles comes
+  after it. A step that installs anything else as well, such as a
+  `script/bootstrap` that also installs node packages, comes after it too. A new
+  value reruns every `RUN` step after that line; `COPY` steps are compared by
+  content and need nothing. Docker keeps the layers above the line until the
+  base image, their own lines or the files they copy change. A stage without the
+  `ARG` line is served from cache on an unchanged tree, so a repository
+  re-vendors the scripts and adds the `ARG` lines in one commit. The scripts do
+  not pass `--no-cache`: it reruns every step too, but it also builds the layers
+  above the line again and empties every cache mount, so Go's build cache and
+  golangci-lint's, which the canonical Go `Dockerfile` below mounts so that they
+  survive between check builds, would start empty every time. A bare
+  `docker build .` is not evidence that anything ran: a sub-second build
+  reporting success is a cache hit, not a result. Never invalidate by pruning —
+  `docker builder prune` and friends destroy a build cache shared with every
+  other build on the host. When a check is added or changed, prove it works by
+  planting a defect it must catch and watching the run fail on it, then revert
+  the defect. A green run alone shows neither that the check ran nor that it
+  covers what it should.
+
+- **The gate phases are separate stages, and the build stage depends on both.**
+  The lint phase is based on the `golangci/golangci-lint` image (pinned by
+  hash), so lint failures surface in seconds rather than after a full compile,
+  and the test phase is based on the Debian Go image. The canonical Go repo
+  `Dockerfile`:
 
     ```dockerfile
-    # Lint stage — fast feedback on formatting and lint issues
+    # In every stage ARG CACHEBUST comes right after the steps that install
+    # system packages and download Go's modules, or right after FROM in a
+    # stage with neither, and every step that lints, tests or compiles
+    # comes after it. script/ passes a new value on every build, which
+    # reruns every RUN step after it. Docker keeps the layers above it
+    # until the base image, their own lines or go.mod and go.sum change.
+    # The steps that compile mount Go's build cache, and golangci-lint run
+    # also mounts golangci-lint's; both survive from build to build.
+
+    # Lint phase
     # golangci/golangci-lint:v2.x.x, YYYY-MM-DD
     FROM golangci/golangci-lint@sha256:... AS lint
     WORKDIR /src
     COPY go.mod go.sum ./
     RUN go mod download
+    ARG CACHEBUST
     COPY . .
-    RUN make fmt-check
+    RUN --mount=type=cache,target=/root/.cache/go-build \
-    RUN make lint
+        --mount=type=cache,target=/root/.cache/golangci-lint \
+        golangci-lint run --config .golangci.yml ./...
 
-    # Build stage
+    # Test phase. -race needs cgo and so a C compiler, which the Debian Go
-    # golang:1.x-alpine, YYYY-MM-DD
+    # image ships and the alpine one does not. -count=1 because the mounted
-    FROM golang@sha256:... AS builder
+    # cache also holds test results, which go test would otherwise replay.
+    # golang:1.x, YYYY-MM-DD
+    FROM golang@sha256:... AS test
     WORKDIR /src
-
-    # Force BuildKit to run the lint stage before proceeding
-    COPY --from=lint /src/go.sum /dev/null
-
     COPY go.mod go.sum ./
     RUN go mod download
+    ARG CACHEBUST
     COPY . .
-    RUN make test
+    RUN --mount=type=cache,target=/root/.cache/go-build \
+        go test -count=1 -timeout 90s -race -cover ./... || \
+        { echo "--- Rerunning with -v for details ---"; \
+          go test -count=1 -timeout 90s -race -v ./...; exit 1; }
 
-    ARG VERSION=dev
+    # Build stage. Nothing is wanted from either phase above; the copies
-    RUN CGO_ENABLED=0 go build -trimpath \
+    # are what make BuildKit build them first, so nothing after them can
+    # run unless lint and test passed.
+    # golang:1.x-alpine, YYYY-MM-DD
+    FROM golang@sha256:... AS builder
+    RUN apk add --no-cache git
+    WORKDIR /src
+    COPY go.mod go.sum ./
+    RUN go mod download
+    ARG CACHEBUST
+    COPY --from=lint /src/go.sum /dev/null
+    COPY --from=test /src/go.sum /dev/null
+    # A tar-stream context keeps the sender's file owners, which git refuses.
+    RUN git config --system --add safe.directory /src
+    COPY . .
+
+    # The VERSION build arg when one is given, otherwise
+    # `git describe --tags --always` on the .git in the build context. With
+    # .git present, a version that is still empty, dev or unknown fails the
+    # build: git is missing or could not read the checkout.
+    ARG VERSION
+    RUN --mount=type=cache,target=/root/.cache/go-build \
+        VERSION="${VERSION:-$(git describe --tags --always)}"; \
+        if [ -e .git ]; then \
+            case "$VERSION" in ""|dev|unknown) \
+                echo "version is '$VERSION' although .git is present" >&2; \
+                exit 1 ;; \
+            esac; \
+        fi; \
+        CGO_ENABLED=0 go build -trimpath \
             -ldflags="-s -w -X main.Version=${VERSION}" \
             -o /app ./cmd/app/
 
-    # Runtime stage
+    # Runtime stage, and the last one
     FROM alpine@sha256:...
+    ARG CACHEBUST
     COPY --from=builder /app /usr/local/bin/app
     ENTRYPOINT ["app"]
     ```
 
     Key points:
-    - The lint stage uses the `golangci/golangci-lint` image directly (it
+    - The lint phase uses the `golangci/golangci-lint` image directly (it has
-      includes both Go and the linter), so there is no need to install the
+      both Go and the linter), so nothing needs installing.
-      linter separately.
+    - `COPY --from=<phase> /src/go.sum /dev/null` is a no-op copy whose only
-    - `COPY --from=lint /src/go.sum /dev/null` is a no-op file copy that creates
+      purpose is the ordering edge. BuildKit runs stages in parallel by default,
-      a stage dependency. BuildKit runs stages in parallel by default; without
+      and a stage nothing depends on is not built at all, so without these two
-      this line, the build stage would not wait for lint to finish and a lint
+      lines a red gate would not fail the build.
-      failure might not fail the overall build.
+    - Keep the runtime stage last, and if you add a stage after it, give it the
+      same two copies. A plain `docker build .` builds the last stage's chain
+      and nothing else.
+    - Check builds keep these, and none of them can hand back a stale result:
+        - The layers above `ARG CACHEBUST`: the system packages and Go's
+          modules. Docker installs the packages again when the base image or the
+          install line changes, and downloads the modules again when `go.mod` or
+          `go.sum` changes. Go checks every module against `go.sum` as it
+          downloads it, and no later build can change a kept layer. The modules
+          are kept in a layer rather than a cache mount because a cache mount is
+          shared by every build on the host, and Go builds from whatever files
+          it holds without checking them again (only `go mod verify` does), so a
+          file changed there would go unnoticed.
+        - Go's build cache (`/root/.cache/go-build`, where Go keeps it for
+          root), mounted on the steps that compile: `golangci-lint run`, both
+          `go test` runs and `go build`. Go reuses a compiled package only when
+          all its inputs are unchanged. The cache also holds test results, which
+          is why both `go test` runs carry `-count=1`.
+        - golangci-lint's cache (`/root/.cache/golangci-lint`, where the pinned
+          image keeps it), mounted on `golangci-lint run`. golangci-lint reuses
+          a package's findings only when the package's files and those of every
+          package it imports, `go.mod`, the linters and their settings in
+          `.golangci.yml`, and the golangci-lint version are all unchanged.
     - If the project uses `//go:embed` directives that reference build artifacts
-      (e.g. a web frontend compiled in a separate stage), the lint stage must
+      (e.g. a web frontend compiled in a separate stage), the lint phase must
       create placeholder files so the embed directives resolve. Example:
       `RUN mkdir -p web/dist && touch web/dist/index.html web/dist/style.css`.
-      The lint stage should not depend on the actual build output — it exists to
+    - If the project requires CGO or system libraries for linting, install them
-      fail fast.
+      in the lint phase. The `golangci/golangci-lint` image is Debian-based and
-    - If the project requires CGO or system libraries for linting (e.g.
+      has no `apk`, so install with `apt-get` under the Debian package name
-      `vips-dev`), install them in the lint stage with `apk add`.
+      (`libvips-dev`, where alpine says `vips-dev`), and delete the package
-    - The build stage runs `make test` after compilation setup. Tests run in the
+      lists in the same `RUN`, so the layer does not keep them. The test phase,
-      build stage, not the lint stage, because they may require compiled
+      on the Debian Go image, installs the libraries its tests need the same
-      artifacts or heavier dependencies.
+      way. Either install goes right after `FROM`, above `ARG CACHEBUST`, so
+      Docker keeps it until the base image or the install line changes:
+
+        ```dockerfile
+        FROM golangci/golangci-lint@sha256:... AS lint
+        RUN apt-get update \
+            && apt-get install -y --no-install-recommends libvips-dev \
+            && rm -rf /var/lib/apt/lists/*
+        WORKDIR /src
+        COPY go.mod go.sum ./
+        RUN go mod download
+        ARG CACHEBUST
+        ```
+
+    - `.dockerignore` lets `.git` into the build context. It keeps out every git
+      `config` at any depth (`**/.git/config`, `**/.git/modules/**/config`): the
+      repository's own, each submodule's under `.git/modules/`, and that of a
+      submodule keeping its own `.git` directory. `git describe` does not need
+      them, and each can hold a credential: a password in a remote URL, or the
+      token the CI checkout step stores there. A submodule whose name has a
+      `config` segment (`config`, `deploy/config`, `config/lib`) loses its whole
+      git directory to `**/.git/modules/**/config`, and Go's version stamping
+      then fails the build: give it a name without that segment
+      (`git submodule add --name`). The stage that compiles has `git` (the
+      Debian Go image has it; an alpine one needs `apk add --no-cache git`) and
+      takes the version from the `VERSION` build argument when one is given,
+      otherwise from `git describe --tags --always`. That gives the tag on a
+      tagged commit; on a later commit, the tag, the number of commits since it
+      and the short commit (`v1.2.3-4-gabc1234`); and the short commit when no
+      tag is reachable. The stage that compiles also marks its working directory
+      safe for git (`git config --system --add safe.directory /src`): a context
+      sent as a tar stream keeps the sender's file owners, and git refuses a
+      checkout owned by another user, so the version would come out empty.
+      `ARG VERSION` has no default, and the build fails if the context carries
+      `.git` and the version still comes out empty, `dev` or `unknown`. A plain
+      `docker build .` with no build arguments must succeed; a Dockerfile that
+      refuses an empty build argument drops that refusal and keeps the argument.
+      A checkout whose `.git` is a file (a linked worktree, or a repository
+      checked out as a submodule) is the exception: that file points to a git
+      directory outside the build context, so the build cannot read the version
+      and a plain `docker build .` fails; pass the version with
+      `--build-arg VERSION=...`, as `script/docker` and `script/cibuild` already
+      do.
 
 - Every repo should have a Gitea Actions workflow (`.gitea/workflows/`) that
-  runs `script/cibuild` (which runs `docker build .`) on push. Since the
+  runs `script/cibuild` on push, and checks out the repo as its only other step,
-  Dockerfile already runs `make check`, a successful build implies all checks
+  with `persist-credentials: false`: `script/cibuild` needs no token, and
-  pass.
+  without it the checkout leaves the job's token in `.git/config` for every
+  later step. The checkout step also sets `fetch-depth: 0`, which fetches the
+  tags `git describe` needs: by default it clones shallow with no tags, and a
+  tagged repository's CI build would stamp a bare short commit id. The
+  workflow's `concurrency` block groups runs by workflow and branch
+  (`${{ github.workflow }}-${{ github.ref }}`) with `cancel-in-progress: true`,
+  so a new push cancels the older run on the same branch, queued or running, and
+  no other: runs for replaced commits do not hold up the shared runner.
+  `script/cibuild` bootstraps, runs the gate phases, and then builds the image,
+  so a successful run means every check passed; a bare `docker build .` does not
+  carry the same guarantee, because its gate phases may come from the cache. The
+  image build gets a `CACHEBUST` value of its own and so runs the gate phases a
+  second time. That is the price of the rule above, and it is worth paying: the
+  image that ships is built from a run of its own gates rather than from a cache
+  entry. The `check` job sets `timeout-minutes: 20`, so a hung build frees the
+  shared runner after 20 minutes. That allows for the three Docker builds
+  described above (the test phase, the lint phase, then the image), each held to
+  the 5-minute Docker build limit below, plus the bootstrap. A separate workflow
+  limited to `main` by a `branches` list under `on: push` cannot be checked by
+  review: to try a change to it, add the feature branch to that list and push,
+  then remove the branch from the list again before merging. Keep any job in it
+  that publishes behind `if: github.ref_name == 'main'`, so the run from the
+  feature branch publishes nothing.
 
 - Use platform-standard formatters: `black` for Python, `prettier` for
   JS/CSS/Markdown/HTML, `go fmt` for Go. Always use default configuration with
@@ -189,14 +408,21 @@ style conventions are in separate documents:
   module under test to verify it compiles/parses. There is no excuse for
   `make test` to be a no-op.
 
-- `make test` must complete in under 20 seconds. Add a 30-second timeout in the
+- `make test` must complete in under 60 seconds. That is the hard cap, and a
-  Makefile.
+  suite that exceeds it fails. Under 20 seconds is the target. A suite between
+  20 and 60 seconds is still green, but the overage must be filed as an
+  improvement bug against that repo. Add a 90-second timeout to the test
+  invocation (`go test -timeout 90s`). The backstop deliberately sits above the
+  hard cap so that it catches a genuinely hung test rather than a merely slow
+  one.
 
-- **`make test` should use the conditional verbose rerun pattern.** Run tests
+- **The test command should use the conditional verbose rerun pattern.** Run
-  without `-v` (verbose) first. If tests fail, automatically rerun with `-v` to
+  tests without `-v` (verbose) first. If tests fail, automatically rerun with
-  show full output. This keeps CI logs and `docker build` output clean on
+  `-v` to show full output. This keeps CI logs and `docker build` output clean
-  success (just package/suite summaries) while providing full diagnostic detail
+  on success (just package/suite summaries) while providing full diagnostic
-  on failure (every test case, every assertion). The general shell pattern:
+  detail on failure (every test case, every assertion). The command lives in the
+  `test` phase of the `Dockerfile`, since `script/test` builds that phase; the
+  Makefile form below is the same pattern for any repo-local invocation:
 
     ```makefile
     test:
@@ -209,11 +435,25 @@ style conventions are in separate documents:
 
     ```makefile
     test:
-    	@go test -timeout 30s -race -cover ./... || \
+    	@go test -count=1 -timeout 90s -race -cover ./... || \
     		{ echo "--- Rerunning with -v for details ---"; \
-    		  go test -timeout 30s -race -v ./...; exit 1; }
+    		  go test -count=1 -timeout 90s -race -v ./...; exit 1; }
     ```
 
+    `-count=1` is required on both invocations: it defeats Go's test _result_
+    cache, so neither run can report a stored pass in place of running the
+    tests. It leaves the build cache alone, so it costs the runtime of the suite
+    and no recompilation.
+
+    That cache is Go's own, separate from Docker's layer cache. Go stores a
+    passing result in its cache directory (`GOCACHE`), and when the same tests
+    run again on unchanged code it prints that result, marked `(cached)`,
+    without running them. The directory lasts from one run to the next on a
+    developer's machine, where this target runs, and in the `test` phase of the
+    `Dockerfile`, which mounts it to keep compiled packages between check
+    builds, so the phase's `go test` runs carry `-count=1` too. A new
+    `CACHEBUST` value (above) is what makes that step run on an unchanged tree.
+
     Python example:
 
     ```makefile
@@ -239,10 +479,89 @@ style conventions are in separate documents:
   must be in `.gitignore`. No exceptions.
 
 - `.gitignore` should be comprehensive from the start: OS files (`.DS_Store`),
-  editor files (`.swp`, `*~`), language build artifacts, and `node_modules/`.
+  editor files (`.swp`, `*~`), in-repo agent scratch directories (`.claude/`),
-  Fetch the standard `.gitignore` from
+  `node_modules/`, and the repo's own build outputs. Fetch the standard
+  `.gitignore` from
   `https://git.eeqj.de/sneak/prompts/raw/branch/main/.gitignore` when setting up
-  a new repo.
+  a new repo. A repo's `.gitignore` is the standard file followed by the repo's
+  own entries, such as its binaries; a re-vendor replaces the standard part and
+  keeps those entries. These patterns are written to `.gitignore`'s own
+  semantics, in which an unanchored pattern already matches at every depth; they
+  are not a `.dockerignore` and must not be transplanted into one unmodified.
+
+- **`.dockerignore` does not use `.gitignore` semantics, and copying patterns
+  across unmodified leaves secrets in the build context.** Docker matches with
+  `moby/patternmatcher`: `filepath.Match` semantics plus a `**` extension, so
+  `*` does not cross `/` and a pattern without a leading `**/` is anchored at
+  the build-context root. A `.dockerignore` listing `.env`, `*.pem` and `*.key`
+  therefore excludes only the copies at the repository root, while `config/.env`
+  and `certs/server.key` still reach the context and can land in an image layer
+  — which is more dangerous than a short file with no secret patterns at all,
+  because it reads as solved and stops anyone looking. Give every
+  depth-independent pattern the `**/` prefix and leave only genuinely
+  root-anchored entries unprefixed: `.claude`, and the repo's own host-built
+  binary, written `/myapp` and never `**/myapp`, which would also match
+  `cmd/myapp/` and delete the package directory from the context. Matching is
+  case-sensitive, and an ALL-CAPS twin per pattern still misses `Server.Key`, so
+  secret names use character ranges — `**/*.[kK][eE][yY]`, `**/*.[pP][eE][mM]`,
+  and likewise for `.envrc` and the extensionless SSH keys. Where such a pattern
+  also catches something the build needs, re-include it with a negation
+  (`!docs/example.env`); deleting the pattern reopens the exposure for every
+  other file it covers. Fetch the standard `.dockerignore` from
+  `https://git.eeqj.de/sneak/prompts/raw/branch/main/.dockerignore` and extend
+  it with the repo's own artifacts.
+
+- **In-repo agent scratch belongs in both files, written to each file's own
+  semantics.** `.claude/` holds one worktree per in-flight agent — an entire
+  additional checkout of the repo — so under `COPY . .` the build context
+  inflates by a multiple of the repo and another session's unreviewed work can
+  be copied into an image layer. In `.gitignore` the entry is `.claude/`,
+  unanchored. In `.dockerignore` it is `.claude`, anchored and with **no** `**/`
+  prefix, because the prefixed form would also delete any nested directory of
+  that name from the build. Anchoring carries a known gap that the canonical
+  `.dockerignore` states in its own comment, since consuming repos receive the
+  file and not the tracker: the directory is created in the agent's working
+  directory, so a repo running agents in subdirectories still ships
+  `services/api/.claude/` and must add its own anchored entry there.
+
+- **A plain `docker build .` of a clone stamps the version that
+  `git describe --tags --always` gives**, derived from the `.git` in the build
+  context as the canonical `Dockerfile` above shows. Without its failure check,
+  a missing `git` or an unreadable checkout would leave `-X main.Version=` empty
+  and the build would still exit 0. `script/docker` and `script/cibuild` pass
+  the version they compute on the host; it takes precedence. They do this
+  byte-identically across repos:
+
+    ```sh
+    # The version and the tag each get their own line: a failing command
+    # substitution inside an argument does not trip `set -e`, so the inline
+    # form degrades to an empty constant.
+    version="$(git describe --tags --always --dirty 2>/dev/null || true)"
+    [ -n "$version" ] || version="unknown"
+    tag="$(script/projectname)"
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
+        --build-arg VERSION="$version" \
+        -t "$tag" .
+    ```
+
+    `--always` makes an untagged repo yield an abbreviated commit hash rather
+    than failing, and the `[ -n "$version" ]` line is the single place the
+    fallback is applied — a live check that fires on a build from an export with
+    no `.git` and on a repository with no commits yet. Do not fold it into the
+    substitution as `|| echo unknown`, which makes the guard unreachable. The
+    Dockerfile's side is `ARG VERSION` in the stage that compiles, declared
+    there because `ARG` is stage-scoped; passing `VERSION` to a repo whose
+    Dockerfile declares no such `ARG` is ignored and costs nothing, which is why
+    the scripts stay byte-identical. One consequence for CI: the standard
+    checkout action clones shallow and fetches no tags, so the canonical
+    `.gitea/workflows/check.yml` sets `fetch-depth: 0` on its checkout step.
+
+- **Verify `.dockerignore` by enumerating the image, not by reading the
+  patterns.** Plant files at the root _and_ at least two directories deep, build
+  a probe image that does `COPY . .`, and list what actually landed
+  (`docker run --rm --entrypoint find IMAGE /app`). The `transferring context`
+  size is not a substitute: a nested secret is a few bytes, and BuildKit
+  transfers only the delta from the previous build.
 
 - **No build artifacts in version control.** Code-derived data (compiled
   bundles, minified output, generated assets) must never be committed to the
@@ -258,9 +577,61 @@ style conventions are in separate documents:
 - Make all changes on a feature branch. You can do whatever you want on a
   feature branch.
 
-- `.golangci.yml` is standardized and must _NEVER_ be modified by an agent, only
+- `.golangci.yml` is standardized. The vendored copy in a consuming repo must
-  manually by the user. Fetch from
+  _NEVER_ be modified by an agent: fetch it from
-  `https://git.eeqj.de/sneak/prompts/raw/branch/main/.golangci.yml`.
+  `https://git.eeqj.de/sneak/prompts/raw/branch/main/.golangci.yml` and keep it
+  byte-identical, so that no repo can quietly loosen its own linting. Linter
+  configuration changes are made to the canonical copy in the `prompts` repo and
+  reach consuming repos by re-vendoring; an agent may open a PR against
+  canonical, which only the user merges. One list is exempt from byte-identity,
+  because it cannot be written once for every repo: the `deny` list of the
+  `test-support` depguard rule, where a repo names its own test-support packages
+  by full import path. A repo adds entries there and changes nothing else, and a
+  re-vendor carries its entries forward. The canonical golangci-lint version is
+  v2.14.0 (released 2026-09-24), pinned as the digest of the lint phase's base
+  image
+  (`golangci/golangci-lint@sha256:ad862ba6b3798cbe0fd9fd7408d498fd74fbd2623a92406b2fd3898faf0bf98f`,
+  which reports `2.14.0 built with go1.27.0 from 114493f9`). A module's `go`
+  directive must not name a newer Go minor version than the one golangci-lint
+  was built with, or golangci-lint refuses to lint it: this release lints
+  `go 1.27.1` but not `go 1.28`. That digest is the only pin, since no repo
+  installs golangci-lint on the host. A repo sets the lint phase digest to the
+  one named here and re-vendors `.golangci.yml` in the same commit, whichever of
+  the two prompted the change: the canonical copy can name linters that an older
+  golangci-lint rejects, and a newer golangci-lint can add linters that
+  `default: all` switches on until the canonical copy disables them.
+
+- **`script/bootstrap` installs a pinned tool by comparing versions, never by
+  testing presence.** An `if ! command -v <tool>; then install; fi` guard tests
+  `PATH` only, so on an already-provisioned machine the pin is inert and a
+  version bump is a silent no-op — while the Dockerfile, installing into a clean
+  image, gets the pinned version, so a local `make check` and `make docker` can
+  disagree about what the tool even is. The canonical form:
+    - compares the installed version against the pin over the **whole** version
+      token; a parser that stops at the first `-` reports `2.12.2` for a host
+      running `2.12.2-rc1` and skips the install;
+    - treats absent, non-zero, empty or unrecognised `--version` output as a
+      mismatch, so the failure direction is a redundant install and never a
+      skipped one;
+    - after installing, re-resolves the binary the way callers do — `hash -r`,
+      then through `PATH`, not through the directory the installer wrote to —
+      and fails naming the resolved path, since an install that a shadowing
+      binary hides succeeds while changing nothing any caller sees;
+    - is actually called, and prints the version on both success paths: a
+      function defined and never invoked has the same exit status and the same
+      empty output as one that worked.
+
+    Node is the exception to the whole-token comparison: the canonical
+    `script/bootstrap` compares only its major version, because the node 22
+    alpine image the `Dockerfile` stages start from is not the exact pin, and
+    nvm cannot install a prebuilt node on alpine.
+
+    Keep it POSIX sh: no arrays, no `[[`, no `grep -P`.
+
+    A Go tool a repo needs on the host is installed with `go install` pinned to
+    a commit hash (`go install <package>@<commit hash>`). It is never tracked as
+    a `go.mod` tool dependency or through a `tools.go` file, either of which
+    pulls the tool's own dependencies into the repo's `go.mod` and `go.sum`.
 
 - When pinning images or packages by hash, add a comment above the reference
   with the version and date (YYYY-MM-DD).
@@ -371,15 +742,21 @@ style conventions are in separate documents:
       Never edit existing migrations after release.
 
 - All repos should have an `.editorconfig` enforcing the project's indentation
-  settings.
+  settings: the standard file from
+  `https://git.eeqj.de/sneak/prompts/raw/branch/main/.editorconfig`, which sets
+  tabs for `Makefile` and Go files, followed by the repo's own sections, such as
+  one for another language it uses. A re-vendor replaces the standard part and
+  keeps those sections.
 
 - Avoid putting files in the repo root unless necessary. Root should contain
-  only project-level config files (`README.md`, `Makefile`, `Dockerfile`,
+  only project-level config files (`README.md`, `AGENTS.md`, `Makefile`,
-  `LICENSE`, `.gitignore`, `.editorconfig`, `REPO_POLICIES.md`, and
+  `Dockerfile`, `LICENSE`, `.gitignore`, `.editorconfig`, `REPO_POLICIES.md`,
-  language-specific config). Everything else goes in a subdirectory. Canonical
+  and language-specific config). Everything else goes in a subdirectory.
-  subdirectory names:
+  Canonical subdirectory names:
     - `bin/` — executable scripts and tools
-    - `cmd/` — Go command entrypoints
+    - `cmd/` — Go command entrypoints; thin only: one `main.go` per binary whose
+      body is a single call into `internal/` or `pkg/`, no project logic in
+      `cmd/`
     - `configs/` — configuration templates and examples
     - `deploy/` — deployment manifests (k8s, compose, terraform)
     - `docs/` — documentation and markdown (README.md stays in root)
@@ -406,3 +783,7 @@ style conventions are in separate documents:
     - Go: `go.mod`, `go.sum`, `.golangci.yml`
     - JS: `package.json`, `yarn.lock`, `.prettierrc`, `.prettierignore`
     - Python: `pyproject.toml`
+
+- Guidance for coding agents lives in one `AGENTS.md` at the repository root. It
+  is never committed under a file or directory named after one agent tool, such
+  as `CLAUDE.md` or `.claude/`, and never split into separate memory files.

TODO.md

@@ -1,122 +1,264 @@
 # Workflow
 
-* branch (from `main`)
+- branch (from `main`)
-* do the work in Next Step
+- do the work in Next Step
-* move Next Step to the top of Completed Steps
+- move Next Step to the top of Completed Steps
-* move the top item of Future Steps into Next Step
+- move the top item of Future Steps into Next Step
-* commit (`TODO.md` changes in the same commit as the work)
+- commit (`TODO.md` changes in the same commit as the work)
-* merge to `main` if the branch is not protected, otherwise open a PR
+- merge to `main` if the branch is not protected, otherwise open a PR
-* push
+- push
 
 # Status
 
-pre-1.0. No git tags. The Docker build runs the format check, the linter
+pre-1.0. No git tags. `main` last took `next` on 2026-10-06, through
-and the tests, and the Gitea workflow runs that build on every push. The
+https://git.eeqj.de/sneak/routewatch/pulls/6. The linter and the tests are
-image sets memory ceilings for a 5 GiB container (README "Memory") and the
+phases of the Docker build, and the Gitea workflow runs `script/cibuild` (the
-README says how to run it under upaas (README "Running under upaas"). A
+format check of the Go code and the markdown, both phases, then the image) on
-35-hour run of `3898daa` on the live feed peaked at about 1 GiB, without a
+every push. The image sets memory ceilings for a 5 GiB container (README
-container memory limit.
+"Memory") and the README says how to run it under upaas (README "Running under
+upaas"). A 25-hour run of `a9b6860` on the live feed peaked at about 0.6 GiB
+(https://git.eeqj.de/sneak/routewatch/issues/3). A run of `825afa2` on the live
+feed found the prefix handler dropping about a quarter of its updates because
+its route writes were slower than the feed
+(https://git.eeqj.de/sneak/routewatch/issues/96); the cheaper writes since then
+are measured by a benchmark, not yet on the live feed.
 
 # Next Step
 
-`next` waits for sneak to merge it to `main` through
+Run `next` on the live feed, as README "Running under upaas" describes, for the
-https://git.eeqj.de/sneak/routewatch/pulls/6. After that, setting
+final check https://git.eeqj.de/sneak/routewatch/issues/96 names: whether the
-routewatch up under upaas on fsn1app1 and deploying it are his
+prefix handler still drops updates after the first hour. If it does, a new issue
-(https://git.eeqj.de/sneak/routewatch/issues/31), and so is the run under a
+follows with the numbers. Then an audit of the code at the commit that run
-real 5 GiB limit (https://git.eeqj.de/sneak/routewatch/issues/3).
+tested.
 
 # Completed Steps
 
-- 2026-10-03: `/api/v1/stats` serves the prefix distribution from memory,
+- 2026-10-27: the prefix handler applies each update to the routing table in
-  seeded at startup and adjusted on every live-route write, so a request no
+  memory, and writes the table's changes to the database every second. A write
-  longer reads every live route (closes #30)
+  takes at most 20,000 changed pairs, and the rest follow in the next writes. A
-- 2026-10-03: a new database is created with `auto_vacuum` set to
+  failed write is written again later. Startup reads the table before the
-  incremental, through the connection string so it is set before the file
+  streamer connects. The pages still read the per-peer route tables (part of
-  is first written, and each periodic incremental vacuum now returns up to
+  #122)
-  1000 free pages; the late `PRAGMA auto_vacuum` in `Initialize`, which
+- 2026-10-27: the database keeps a copy of the whole routing table in a new
-  SQLite ignored, is gone (closes #43)
+  `routes` table, one row per prefix and origin AS. The peers' numbers are in a
-- 2026-10-03: looking up an IP address no longer reads every IPv6 route:
+  new `route_peers` table. One function writes the pairs that changed in memory,
-  both families find the most specific live route with at most 33 or 129
+  and another reads the whole table back into memory at startup. Nothing uses
-  lookups on the prefix index, and the IPv4 range columns are gone. Prefixes
+  them yet (part of #122)
-  from the feed are stored in one text form, so an IPv6 withdrawal, which the
+- 2026-10-27: a new package, `internal/routetable`, holds the whole routing
-  feed sends uncompressed, now removes its route (closes #48)
+  table in memory. For each prefix and origin AS it keeps the peers that carry
-- 2026-10-02: a plain `docker build .` stamps the commit's tag or short
+  it, one bit per peer, and the two newest of their routes. Nothing uses it yet
-  commit (`git describe --tags --always`) into the page footer instead of
+  (part of #122)
-  `unknown`: `.dockerignore` sends `.git` without `.git/config`, a `VERSION`
+- 2026-10-27: each result of the IP lookup lists the public blocklists the
-  build argument takes precedence, and `make build` stamps the same value
+  address is on, in `blocklists`. The lists are DShield, Feodo Tracker, two
-  (closes #46)
+  FireHOL lists, Spamhaus DROP and the Tor exit list. RouteWatch downloads each
+  list on its own schedule and answers from copies in memory. The copies are
+  also kept in the state directory's `blocklists` directory. IP lookup answers
+  carry `Cache-Control`. An address with no live route now has a result too,
+  without route fields, so the IP lookup no longer answers 404 (closes #158)
+- 2026-10-27: the home page has an IP information API card. It shows the `curl`
+  command that looks up the visitor's own address at `/ip/{addr}`, and that
+  command's live answer (closes #156)
+- 2026-10-27: on the AS page, a click on the heading of the IPv4 or the IPv6
+  prefix list folds or unfolds the list. Both lists start open (closes #155)
+- 2026-10-27: the AS page lists its IPv4 and its IPv6 prefixes by mask length,
+  shortest first, and prefixes of one mask length by address. They used to be
+  listed by address (closes #153)
+- 2026-10-27: on `/` and `/status`, a spinner beside the status turns one frame
+  each time the stats show more messages than the time before. A page that has
+  not seen more messages for three refreshes shows "stale" and the time since
+  then instead, in red (closes #152)
+- 2026-10-27: the WHOIS details of a Brazilian AS keep their accented letters.
+  LACNIC's answer for such an AS comes from NIC.br, Brazil's registry, with each
+  accented letter replaced. The details now come from NIC.br's own server,
+  `whois.registro.br`. When that query fails, the details from LACNIC's answer
+  are kept and a warning is logged. After NIC.br answers that RouteWatch is over
+  its query limit, RouteWatch stops asking NIC.br for an hour (closes #174)
+- 2026-10-27: `make asupdate` stops with an error when the download of `as.csv`
+  breaks off. The generator used to retry the failed read forever. A record that
+  does not parse is still logged and skipped (closes #177)
+- 2026-10-27: `make asupdate` runs the new `script/asupdate`. It replaces the AS
+  list built into RouteWatch only when the generator succeeds and finds at least
+  one AS. A failed download used to replace the list with an empty file. The
+  generator reads `as.csv` from `ipverse/as-metadata`, the source's new name
+  (closes #173)
+- 2026-10-27: the WHOIS details of an AS at RIPE, APNIC or AFRINIC come only
+  from its `aut-num` object and the organisation object it names. RIPE and APNIC
+  start their answer with the `as-block` object of the AS's block. Its dates
+  were stored as the AS's (closes #168)
+- 2026-10-27: AS names keep their accented letters. The AS list built into
+  RouteWatch is rebuilt from today's `as.csv`, which added a country code
+  column. That file no longer lists AS 0 or the documentation AS numbers. A
+  WHOIS answer that is not valid UTF-8 is decoded as Latin-1, which RIPE and
+  APNIC send, before it is parsed and stored (closes #165)
+- 2026-10-27: the Age column on the AS and prefix pages reads in words, such as
+  "5 minutes ago" or "2 weeks ago", instead of Go durations such as "360h0m0s
+  ago". An age under a minute reads in seconds, and one over a month in months
+  or years instead of a date (closes #163)
+- 2026-10-26: a write to a stored AS, peering or peer keeps the earlier of the
+  two first-seen times and the later of the two last-seen times. An old update
+  from a collector catching up on a backlog no longer moves last seen back
+  (closes #162)
+- 2026-10-26: the status page shows "Route Churn/min", the number of prefix
+  announcements and withdrawals, from all peers, in the last full clock minute.
+  `/api/v1/stats` reports it as `route_churn_per_min` in `stream`, which
+  replaces the moving average `route_churn_per_sec`. The messages and updates
+  per second on `/` and `/status` are whole numbers (closes #160)
+- 2026-10-26: the WHOIS fetcher backs off more slowly. A failure makes its wait
+  1.5 times as long, up to one minute instead of five. Its queries to RIPE,
+  APNIC and AFRINIC leave out person and role objects, which RIPE counts against
+  a daily limit per address. A refused query is logged with the RIR's error line
+  (closes #159)
+- 2026-10-26: the home page's AS name search lists the ASes whose handle or
+  description holds the text, as links to their pages, from the new
+  `/api/v1/search/as`; the search ignores case. When the database has no record
+  of an AS, its page takes the handle and description from the AS list built
+  into RouteWatch and still lists the AS's prefixes (closes #151)
+- 2026-10-19: README "HTTP Endpoints" lists every route with what it answers,
+  "Database Schema" matches `schema.sql`, and the rest of the README matches the
+  code; this file brought up to date (closes #72)
+- 2026-10-19: route writes are cheaper: the live route and prefix tables lose
+  their random UUID keys and the prefix tables their duplicate index, a route
+  update sets only the columns that change, and every write goes through one
+  connection with a page cache of 1 GiB (README "Memory"). A benchmark writes
+  the prefix handler's largest batch faster than the feed sends it (closes #96)
+- 2026-10-19: the WAL checkpoint, the incremental vacuum and ANALYZE take the
+  database lock every batch write holds, so a vacuum no longer fails with
+  "database is locked" and an ANALYZE no longer makes a batch write lose its
+  updates; ANALYZE runs as `PRAGMA optimize` with an analysis limit (closes #94)
+- 2026-10-19: shutdown stops the streamer first, and each handler writes
+  everything left on its queue before it stops, so the updates waiting there are
+  no longer lost (closes #69)
+- 2026-10-19: a path that ends in an AS set of several members has origin AS 0,
+  no known origin, instead of the set's last member, and the set adds no
+  peerings; the pages show such a route as having no known origin (closes #75)
+- 2026-10-19: `bgp_peer_count` drops to 0 when the last peer goes down (closes
+  #92)
+- 2026-10-19: `bgp_peer_count` counts every peer the feed has heard from, not
+  only those whose session opened since startup, and drops a peer whose session
+  goes down (closes #74)
+- 2026-10-19: the shared files are re-vendored from `sneak/prompts` at
+  `0653579`: golangci-lint v2.14.0 and its configuration, lint and test as
+  `Dockerfile` phases that keep Go's and golangci-lint's caches, markdown
+  formatted with prettier, agent guidance in `AGENTS.md`; the code is fixed for
+  what the new linter configuration finds (closes #52)
+- 2026-10-19: the streamer reads feed lines up to 1 MiB, so a line over 64 KiB
+  no longer ends the connection (closes #76)
+- 2026-10-19: a peer's live routes are removed when its BGP session goes down;
+  `STATE` messages are never dropped for backpressure, and one that finds its
+  queue full waits for room ahead of the handler's other messages (closes #56)
+- 2026-10-19: the prefix-length pages pick their sample with a bounded number of
+  index lookups instead of reading every route of that length, and list each
+  prefix once (closes #66)
+- 2026-10-19: a request past its time limit answers 504 instead of 408, and
+  `/api/v1/stats` answers 504 instead of 500 when its read runs out of time
+  (closes #70)
+- 2026-10-19: the AS, peer and prefix handlers write each batch outside the lock
+  their message handling needs, so messages keep coming off their queues during
+  a write; meanwhile the next batch grows to at most ten batch sizes (closes
+  #65)
+- 2026-10-19: a live route is identified by prefix and peer, so an announcement
+  with a new origin AS replaces the peer's route and a withdrawal removes it
+  whatever its origin AS (closes #55)
+- 2026-10-14: the streamer reconnects when RIS Live ends the stream or sends
+  nothing for 60 seconds, and the health check reports the feed disconnected
+  meanwhile (closes #53)
+- 2026-10-14: the prefix page takes the names of the origin ASes from the prefix
+  query instead of reading every route of each origin AS (closes #67)
+- 2026-10-14: the daemon exits with status 1 when its web server cannot listen
+  on its port, instead of running on without it (closes #73)
+- 2026-10-14: the prefix pages parse the prefix before looking it up: one that
+  does not parse answers 400, and any valid way of writing a prefix finds its
+  routes (closes #71)
+- 2026-10-14: the IP lookup ignores routes shorter than /8 for IPv4 and /16 for
+  IPv6, so a default route is no longer named as an address's owner (closes #57)
+- 2026-10-14: a prefix batch keeps each peer's last update for a prefix, instead
+  of one update per prefix across all peers (closes #54)
+- 2026-10-14: `reconnect_count` in `/api/v1/stats` counts every reconnect
+  (closes #68)
+- 2026-10-03: `/api/v1/stats` serves the prefix distribution from memory, seeded
+  at startup and adjusted on every live-route write, so a request no longer
+  reads every live route (closes #30)
+- 2026-10-03: a new database is created with `auto_vacuum` set to incremental,
+  through the connection string so it is set before the file is first written,
+  and each periodic incremental vacuum now returns up to 1000 free pages; the
+  late `PRAGMA auto_vacuum` in `Initialize`, which SQLite ignored, is gone
+  (closes #43)
+- 2026-10-03: looking up an IP address no longer reads every IPv6 route: both
+  families find the most specific live route with at most 33 or 129 lookups on
+  the prefix index, and the IPv4 range columns are gone. Prefixes from the feed
+  are stored in one text form, so an IPv6 withdrawal, which the feed sends
+  uncompressed, now removes its route (closes #48)
+- 2026-10-02: a plain `docker build .` stamps the commit's tag or short commit
+  (`git describe --tags --always`) into the page footer instead of `unknown`:
+  `.dockerignore` sends `.git` without `.git/config`, a `VERSION` build argument
+  takes precedence, and `make build` stamps the same value (closes #46)
 - 2026-09-29: the entrypoint creates the data directory if it is missing and
-  stops the start if a step fails; README "Running under upaas" no longer
+  stops the start if a step fails; README "Running under upaas" no longer asks
-  asks for the host directory to be created first (closes #42)
+  for the host directory to be created first (closes #42)
-- 2026-09-29: `.dockerignore` keeps `.git`, local build output, local
+- 2026-09-29: `.dockerignore` keeps `.git`, local build output, local databases
-  databases and `.env` out of the Docker build context, and so out of the
+  and `.env` out of the Docker build context, and so out of the source archive
-  source archive in the image (closes #39)
+  in the image (closes #39)
-- 2026-09-29: README first line names the MIT license and the author;
+- 2026-09-29: README first line names the MIT license and the author; the
-  the License section now says MIT and links `LICENSE`, and an Author
+  License section now says MIT and links `LICENSE`, and an Author section was
-  section was added; this file brought up to date (closes #38)
+  added; this file brought up to date (closes #38)
 - 2026-09-29: MIT `LICENSE` (closes #1)
-- 2026-09-28: stopping the daemon while the feed is flowing no longer
+- 2026-09-28: stopping the daemon while the feed is flowing no longer panics
-  panics with "send on closed channel": the read loop checks for a stop
+  with "send on closed channel": the read loop checks for a stop just before
-  just before handing a message to the handler queues, and a second
+  handing a message to the handler queues, and a second `Stop` no longer closes
-  `Stop` no longer closes the queues again (closes #34)
+  the queues again (closes #34)
-- 2026-09-28: `docker stop` no longer kills the daemon 2 seconds after the
+- 2026-09-28: `docker stop` no longer kills the daemon 2 seconds after the stop
-  stop signal: the entrypoint switches to the `routewatch` user with
+  signal: the entrypoint switches to the `routewatch` user with `setpriv`
-  `setpriv` instead of `runuser`, so the daemon receives the signal itself
+  instead of `runuser`, so the daemon receives the signal itself and gets the
-  and gets the whole wait `docker stop` allows, up to its own 60-second
+  whole wait `docker stop` allows, up to its own 60-second limit (closes #33)
-  limit (closes #33)
 - 2026-09-28: ready to run under upaas: a set but invalid `PORT`,
-  `XDG_DATA_HOME` or `MALLOC_ARENA_MAX` stops the start, the health
+  `XDG_DATA_HOME` or `MALLOC_ARENA_MAX` stops the start, the health check
-  check follows `PORT`, README "Running under upaas" section (closes
+  follows `PORT`, README "Running under upaas" section (closes #31)
-  #31)
+- 2026-09-22: realtime in-memory database statistics: counts seeded at startup
-- 2026-09-22: realtime in-memory database statistics: counts seeded at
+  and adjusted on every write, oldest/newest route timestamps via index-end
-  startup and adjusted on every write, oldest/newest route timestamps via
+  lookups; `/api/v1/stats` no longer scans the tables (closes #27)
-  index-end lookups; `/api/v1/stats` no longer scans the tables (closes
+- 2026-09-21: batch writes take the write lock when their transaction begins
-  #27)
+  (`_txlock=immediate`), so they wait out a WAL checkpoint instead of failing
-- 2026-09-21: batch writes take the write lock when their transaction
+  with "database is locked" (closes #25)
-  begins (`_txlock=immediate`), so they wait out a WAL checkpoint instead
+- 2026-09-21: `MALLOC_ARENA_MAX=2` in the image caps glibc malloc arenas, so
-  of failing with "database is locked" (closes #25)
+  memory outside the Go runtime no longer grows with the core count (closes #23)
-- 2026-09-21: `MALLOC_ARENA_MAX=2` in the image caps glibc malloc arenas,
+- 2026-09-21: `GOMEMLIMIT=1536MiB` in the image; README Memory section with the
-  so memory outside the Go runtime no longer grows with the core count
+  memory budget and the 5 GiB container limit (closes #13)
-  (closes #23)
+- 2026-09-21: the four handler queues hold at most 20,000 messages each, down
-- 2026-09-21: `GOMEMLIMIT=1536MiB` in the image; README Memory section
+  from 100,000 (closes #11)
-  with the memory budget and the 5 GiB container limit (closes #13)
+- 2026-09-21: two goroutine leaks fixed: the stats handlers after a timeout and
-- 2026-09-21: the four handler queues hold at most 20,000 messages each,
+  the streamer's tickers on every reconnect (closes #12)
-  down from 100,000 (closes #11)
+- 2026-09-21: parsed RIS messages no longer keep the unused `Community` and
-- 2026-09-21: two goroutine leaks fixed: the stats handlers after a
+  `Raw` fields (closes #9)
-  timeout and the streamer's tickers on every reconnect (closes #12)
+- 2026-09-21: `.editorconfig`, and a Gitea workflow that runs `script/cibuild`
-- 2026-09-21: parsed RIS messages no longer keep the unused `Community`
+  on every push (closes #14)
-  and `Raw` fields (closes #9)
+- 2026-09-21: the peering handler's AS-path map holds at most 500,000 paths and
-- 2026-09-21: `.editorconfig`, and a Gitea workflow that runs
+  is swapped for an empty one every 30 seconds instead of copied (closes #10)
-  `script/cibuild` on every push (closes #14)
+- 2026-09-21: SQLite memory bounded across the whole connection pool: a 64 MiB
-- 2026-09-21: the peering handler's AS-path map holds at most 500,000
+  page cache on each connection, 1 GiB soft and 1.5 GiB hard heap limits (closes
-  paths and is swapped for an empty one every 30 seconds instead of copied
+  #8)
-  (closes #10)
+- 2026-09-21: the Docker build runs the format check, the linter and the tests,
-- 2026-09-21: SQLite memory bounded across the whole connection pool: a
+  linting in a separate stage on a golangci-lint image pinned by digest (closes
-  64 MiB page cache on each connection, 1 GiB soft and 1.5 GiB hard heap
+  #5)
-  limits (closes #8)
-- 2026-09-21: the Docker build runs the format check, the linter and the
-  tests, linting in a separate stage on a golangci-lint image pinned by
-  digest (closes #5)
 - 2026-09-21: `make test` skips the live-network feed test (`-short`), so
   `make check` no longer depends on the network (closes #2)
-- 2026-07-07 Adopted scripts-to-rule-them-all: `script/` entrypoints,
+- 2026-07-07 Adopted scripts-to-rule-them-all: `script/` entrypoints, Makefile
-  Makefile shims, README Entrypoints section
+  shims, README Entrypoints section
-- 2026-02-22: repo policy compliance: required policy files, .gitignore
+- 2026-02-22: repo policy compliance: required policy files, .gitignore update,
-  update, Makefile fmt-check/check/docker/hooks targets, gofmt pass
+  Makefile fmt-check/check/docker/hooks targets, gofmt pass
   (repo-policies-compliance, unmerged)
 - 2026-01-01: WAL checkpointing: periodic (5s), on startup with logging,
   TRUNCATE mode; fixed slow queries
-- 2025-12-31: status page: navbar, home page with search, oldest/newest
+- 2025-12-31: status page: navbar, home page with search, oldest/newest route
-  route timestamps, NULL handling in WHOIS stats; container runs as
+  timestamps, NULL handling in WHOIS stats; container runs as routewatch user
-  routewatch user with proper state dir
+  with proper state dir
-- 2025-12-30: structured HTTP request logging, increased timeouts,
+- 2025-12-30: structured HTTP request logging, increased timeouts, CIDR prefix
-  CIDR prefix URL routing fixes, status page metrics and footer
+  URL routing fixes, status page metrics and footer
 - 2025-12-29: Dockerfile multi-stage build with source archive; SQLite
   incremental vacuum for non-blocking space reclamation
 - 2025-07-27: initial RouteWatch BGP stream monitor
 
 # Future Steps
 
-- Production memory under 5 GiB: whether to test under a real 5 GiB
+- the rest of https://git.eeqj.de/sneak/routewatch/issues/122, one PR each: the
-  container limit on fsn1app1 is open for sneak
+  pages and the API reading from the routing table, then removing the per-peer
-  (https://git.eeqj.de/sneak/routewatch/issues/3)
+  route tables

cmd/asinfo-gen/main.go

@@ -2,8 +2,11 @@
 package main
 
 import (
+	"context"
 	"encoding/csv"
 	"encoding/json"
+	"errors"
+	"fmt"
 	"io"
 	"log"
 	"net/http"
@@ -12,7 +15,21 @@ import (
 	"strings"
 )
 
-const asCsvURL = "https://github.com/ipverse/asn-info/raw/refs/heads/master/as.csv"
+const asCsvURL = "https://github.com/ipverse/as-metadata/raw/refs/heads/master/as.csv"
+
+// The first columns of as.csv, the ones the generator reads. It reads past
+// the columns after them, such as the country code.
+const (
+	asnColumn         = "asn"
+	handleColumn      = "handle"
+	descriptionColumn = "description"
+	columnsRead       = 3
+)
+
+var (
+	errUnexpectedHeader = errors.New("unexpected CSV header")
+	errNoASes           = errors.New("no AS in the CSV")
+)
 
 // ASInfo represents information about an Autonomous System
 type ASInfo struct {
@@ -27,49 +44,82 @@ func main() {
 
 	// Fetch the CSV data
 	log.Println("Fetching AS CSV data...")
-	resp, err := http.Get(asCsvURL)
+
+	req, err := http.NewRequestWithContext(
+		context.Background(), http.MethodGet, asCsvURL, nil)
+	if err != nil {
+		log.Fatalf("Failed to fetch CSV: %v", err)
+	}
+
+	resp, err := http.DefaultClient.Do(req)
 	if err != nil {
 		log.Fatalf("Failed to fetch CSV: %v", err)
 	}
-	defer func() {
-		_ = resp.Body.Close()
-	}()
 
 	if resp.StatusCode != http.StatusOK {
 		log.Fatalf("Failed to fetch CSV: HTTP %d", resp.StatusCode)
 	}
 
-	// Parse CSV
+	asInfos, err := parseCSV(resp.Body)
-	reader := csv.NewReader(resp.Body)
+
+	// Closed here rather than deferred: log.Fatalf exits without running
+	// deferred calls.
+	_ = resp.Body.Close()
+
+	if err != nil {
+		log.Fatalf("Failed to parse CSV: %v", err)
+	}
+
+	log.Printf("Parsed %d AS records", len(asInfos))
+
+	// Convert to JSON and output to stdout
+	encoder := json.NewEncoder(os.Stdout)
+	encoder.SetIndent("", "\t")
+
+	err = encoder.Encode(asInfos)
+	if err != nil {
+		log.Fatalf("Failed to encode JSON: %v", err)
+	}
+}
+
+// parseCSV reads as.csv from r and returns its ASes, with the handle and the
+// description as the file spells them. A header whose first columns are not
+// the AS number, the handle and the description is an error. A record that
+// does not parse is logged and skipped. A read that fails, such as on a broken
+// download, is an error. A file left with no AS is an error.
+func parseCSV(r io.Reader) ([]ASInfo, error) {
+	reader := csv.NewReader(r)
 
-	// Read header
 	header, err := reader.Read()
 	if err != nil {
-		log.Fatalf("Failed to read CSV header: %v", err)
+		return nil, fmt.Errorf("read CSV header: %w", err)
 	}
 
-	if len(header) != 3 || header[0] != "asn" || header[1] != "handle" || header[2] != "description" {
+	if len(header) < columnsRead || header[0] != asnColumn ||
-		log.Fatalf("Unexpected CSV header: %v", header)
+		header[1] != handleColumn || header[2] != descriptionColumn {
+		return nil, fmt.Errorf("%w: %v", errUnexpectedHeader, header)
 	}
 
-	// Read all records
 	var asInfos []ASInfo
+
 	for {
+		// The reader returns an error for a record with a different number
+		// of fields than the header.
 		record, err := reader.Read()
-		if err == io.EOF {
+		if errors.Is(err, io.EOF) {
 			break
 		}
-		if err != nil {
+
+		if _, ok := errors.AsType[*csv.ParseError](err); ok {
 			log.Printf("Error reading CSV record: %v", err)
 
 			continue
 		}
 
-		const expectedFields = 3
+		// Any other error comes from the download, such as a broken
-		if len(record) != expectedFields {
+		// connection. Every later read returns it again.
-			log.Printf("Skipping invalid record: %v", record)
+		if err != nil {
-
+			return nil, fmt.Errorf("read CSV record: %w", err)
-			continue
 		}
 
 		asn, err := strconv.Atoi(record[0])
@@ -86,12 +136,9 @@ func main() {
 		})
 	}
 
-	log.Printf("Parsed %d AS records", len(asInfos))
+	if len(asInfos) == 0 {
+		return nil, errNoASes
+	}
 
-	// Convert to JSON and output to stdout
+	return asInfos, nil
-	encoder := json.NewEncoder(os.Stdout)
-	encoder.SetIndent("", "\t")
-	if err := encoder.Encode(asInfos); err != nil {
-		log.Fatalf("Failed to encode JSON: %v", err)
-	}
 }

cmd/asinfo-gen/main_test.go

@@ -0,0 +1,142 @@
+package main
+
+import (
+	"errors"
+	"io"
+	"slices"
+	"strings"
+	"testing"
+	"testing/iotest"
+	"time"
+)
+
+// parseTimeLimit is how long a test waits for parseCSV to return before it
+// fails.
+const parseTimeLimit = 2 * time.Second
+
+// TestParseCSVReadsTodaysLayout parses an as.csv with the country code column
+// that the source added after the description. It expects each AS with its
+// handle and description spelled exactly as in the file, accents included.
+func TestParseCSVReadsTodaysLayout(t *testing.T) {
+	t.Parallel()
+
+	const asCSV = `asn,handle,description,country-code
+1,LVLT-1,Level 3 Parent LLC,US
+1251,FUNDACAO-AMPARO-PESQUISA,FUNDAÇÃO DE AMPARO À PESQUISA DO ESTADO SÃO PAULO,BR
+`
+
+	got, err := parseCSV(strings.NewReader(asCSV))
+	if err != nil {
+		t.Fatalf("parseCSV: %v", err)
+	}
+
+	want := []ASInfo{
+		{ASN: 1, Handle: "LVLT-1", Description: "Level 3 Parent LLC"},
+		{
+			ASN:    1251,
+			Handle: "FUNDACAO-AMPARO-PESQUISA",
+			Description: "FUNDAÇÃO DE AMPARO À PESQUISA DO ESTADO " +
+				"SÃO PAULO",
+		},
+	}
+
+	if !slices.Equal(got, want) {
+		t.Errorf("parseCSV = %+v, want %+v", got, want)
+	}
+}
+
+// TestParseCSVRejectsAnotherHeader parses files whose first columns are not
+// the AS number, the handle and the description. It expects an error rather
+// than a list read from the wrong columns.
+func TestParseCSVRejectsAnotherHeader(t *testing.T) {
+	t.Parallel()
+
+	for _, asCSV := range []string{
+		"asn,handle\n1,LVLT-1\n",
+		"asn,description,handle\n1,Level 3 Parent LLC,LVLT-1\n",
+	} {
+		_, err := parseCSV(strings.NewReader(asCSV))
+		if err == nil {
+			t.Errorf("parseCSV(%q): no error", asCSV)
+		}
+	}
+}
+
+// TestParseCSVRejectsAFileWithNoASes parses files that hold no AS, one with
+// only the header and one whose only record has no valid AS number. It
+// expects errNoASes, so that make asupdate keeps the old list.
+func TestParseCSVRejectsAFileWithNoASes(t *testing.T) {
+	t.Parallel()
+
+	for _, asCSV := range []string{
+		"asn,handle,description,country-code\n",
+		"asn,handle,description,country-code\nAS1,LVLT-1,Level 3,US\n",
+	} {
+		_, err := parseCSV(strings.NewReader(asCSV))
+		if !errors.Is(err, errNoASes) {
+			t.Errorf("parseCSV(%q) = %v, want %v", asCSV, err, errNoASes)
+		}
+	}
+}
+
+// TestParseCSVSkipsARecordThatDoesNotParse parses an as.csv with a record
+// that has too few fields between two valid records. It expects both valid
+// ASes and no error, so that one bad record does not stop make asupdate.
+func TestParseCSVSkipsARecordThatDoesNotParse(t *testing.T) {
+	t.Parallel()
+
+	const asCSV = "asn,handle,description,country-code\n" +
+		"1,LVLT-1,Level 3 Parent LLC,US\n" +
+		"2,UDEL-DCN\n" +
+		"3,MIT-GATEWAYS,Massachusetts Institute of Technology,US\n"
+
+	got, err := parseCSV(strings.NewReader(asCSV))
+	if err != nil {
+		t.Fatalf("parseCSV: %v", err)
+	}
+
+	want := []ASInfo{
+		{ASN: 1, Handle: "LVLT-1", Description: "Level 3 Parent LLC"},
+		{
+			ASN:         3,
+			Handle:      "MIT-GATEWAYS",
+			Description: "Massachusetts Institute of Technology",
+		},
+	}
+
+	if !slices.Equal(got, want) {
+		t.Errorf("parseCSV = %+v, want %+v", got, want)
+	}
+}
+
+// TestParseCSVStopsWhenTheDownloadBreaksOff parses an as.csv whose download
+// breaks off after the first record. From then on every read fails, as on a
+// broken connection. It expects parseCSV to return that error, so that make
+// asupdate keeps the old list.
+func TestParseCSVStopsWhenTheDownloadBreaksOff(t *testing.T) {
+	t.Parallel()
+
+	const asCSV = "asn,handle,description,country-code\n" +
+		"1,LVLT-1,Level 3 Parent LLC,US\n"
+
+	download := io.MultiReader(
+		strings.NewReader(asCSV), iotest.ErrReader(io.ErrUnexpectedEOF))
+
+	// parseCSV runs on its own goroutine, so that a parseCSV that never
+	// returns fails the test instead of hanging it.
+	parsed := make(chan error, 1)
+
+	go func() {
+		_, err := parseCSV(download)
+		parsed <- err
+	}()
+
+	select {
+	case err := <-parsed:
+		if !errors.Is(err, io.ErrUnexpectedEOF) {
+			t.Errorf("parseCSV = %v, want %v", err, io.ErrUnexpectedEOF)
+		}
+	case <-time.After(parseTimeLimit):
+		t.Fatalf("parseCSV did not return within %v", parseTimeLimit)
+	}
+}

cmd/streamdumper/main.go

@@ -26,9 +26,17 @@ func main() {
 
 	// Register raw message handler that prints to stdout
 	s.RegisterRawHandler(func(line string) {
-		fmt.Println(line)
+		_, _ = fmt.Fprintln(os.Stdout, line)
 	})
 
+	// Start streaming before any defer: log.Fatal exits without running
+	// deferred calls.
+	err := s.Start(context.Background())
+	if err != nil {
+		log.Fatal("Failed to start streamer:", err)
+	}
+	defer s.Stop(context.Background())
+
 	// Set up context with cancellation
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
@@ -36,18 +44,13 @@ func main() {
 	// Handle signals
 	sigChan := make(chan os.Signal, 1)
 	signal.Notify(sigChan, os.Interrupt, syscall.SIGTERM)
+
 	go func() {
 		<-sigChan
 		log.Println("Received shutdown signal")
 		cancel()
 	}()
 
-	// Start streaming
-	if err := s.Start(); err != nil {
-		log.Fatal("Failed to start streamer:", err)
-	}
-	defer s.Stop()
-
 	// Wait for context cancellation
 	<-ctx.Done()
 }

go.mod

@@ -1,15 +1,19 @@
 module git.eeqj.de/sneak/routewatch
 
-go 1.24.4
+go 1.27.1
 
 require (
+	github.com/bits-and-blooms/bitset v1.26.0
 	github.com/dustin/go-humanize v1.0.1
 	github.com/go-chi/chi/v5 v5.2.2
+	github.com/google/btree v1.1.3
 	github.com/google/uuid v1.6.0
 	github.com/mattn/go-sqlite3 v1.14.29
 	github.com/rcrowley/go-metrics v0.0.0-20250401214520-65e299d6c5c9
 	go.uber.org/fx v1.24.0
+	go4.org/netipx v0.0.0-20260823151212-3075585bcbeb
 	golang.org/x/term v0.33.0
+	golang.org/x/text v0.43.0
 )
 
 require (

go.sum

@@ -1,9 +1,13 @@
+github.com/bits-and-blooms/bitset v1.26.0 h1:WPvcPZbP1yuKofmjs2SYIFO1EjzIGhCP9ducElLsh5Q=
+github.com/bits-and-blooms/bitset v1.26.0/go.mod h1:7hO7Gc7Pp1vODcmWvKMRA9BNmbv6a/7QIWpPxHddWR8=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/dustin/go-humanize v1.0.1 h1:GzkhY7T5VNhEkwH0PVJgjz+fX1rhBrR7pRT3mDkpeCY=
 github.com/dustin/go-humanize v1.0.1/go.mod h1:Mu1zIs6XwVuF/gI1OepvI0qD18qycQx+mFykh5fBlto=
 github.com/go-chi/chi/v5 v5.2.2 h1:CMwsvRVTbXVytCk1Wd72Zy1LAsAh9GxMmSNWLHCG618=
 github.com/go-chi/chi/v5 v5.2.2/go.mod h1:L2yAIGWB3H+phAw1NxKwWM+7eUH/lU8pOMm5hHcoops=
+github.com/google/btree v1.1.3 h1:CVpQJjYgC4VbzxeGVHfvZrv1ctoYCAI8vbl07Fcxlyg=
+github.com/google/btree v1.1.3/go.mod h1:qOPhT0dTNdNzV6Z/lhRX0YXUafgPLFUh+gZMl761Gm4=
 github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
 github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/mattn/go-sqlite3 v1.14.29 h1:1O6nRLJKvsi1H2Sj0Hzdfojwt8GiGKm+LOfLaBFaouQ=
@@ -24,9 +28,13 @@ go.uber.org/multierr v1.10.0 h1:S0h4aNzvfcFsC3dRF1jLoaov7oRaKqRGC/pUEJ2yvPQ=
 go.uber.org/multierr v1.10.0/go.mod h1:20+QtiLqy0Nd6FdQB9TLXag12DsQkrbs3htMFfDN80Y=
 go.uber.org/zap v1.26.0 h1:sI7k6L95XOKS281NhVKOFCUNIvv9e0w4BF8N3u+tCRo=
 go.uber.org/zap v1.26.0/go.mod h1:dtElttAiwGvoJ/vj4IwHBS/gXsEu/pZ50mUIRWuG0so=
+go4.org/netipx v0.0.0-20260823151212-3075585bcbeb h1:XBM4hvfwGAttkkiTIFfeigdfcL1xIfdKXqFdgiHGtDs=
+go4.org/netipx v0.0.0-20260823151212-3075585bcbeb/go.mod h1:PLyyIXexvUFg3Owu6p/WfdlivPbZJsZdgWZlrGope/Y=
 golang.org/x/sys v0.34.0 h1:H5Y5sJ2L2JRdyv7ROF1he/lPdvFsd0mJHFw2ThKHxLA=
 golang.org/x/sys v0.34.0/go.mod h1:BJP2sWEmIv4KK5OTEluFJCKSidICx8ciO85XgH3Ak8k=
 golang.org/x/term v0.33.0 h1:NuFncQrRcaRvVmgRkvM3j/F00gWIAlcmlB8ACEKmGIg=
 golang.org/x/term v0.33.0/go.mod h1:s18+ql9tYWp1IfpV9DmCtQDDSRBUjKaw9M1eAv5UeF0=
+golang.org/x/text v0.43.0 h1:1QivrlhwAsnMOcqgOdl9my7s9OyZl8O5/UGUTJVXoWc=
+golang.org/x/text v0.43.0/go.mod h1:hbSIYA/amXcRXTFSZqh8tWHEoqRvvbKPJ2uz/KU+a+s=
 gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
 gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=

internal/blocklist/blocklist.go

@@ -0,0 +1,184 @@
+// Package blocklist downloads public IP blocklists on a schedule, keeps the
+// last good copy of each on disk and in memory, and answers which lists
+// contain an address.
+package blocklist
+
+import (
+	"context"
+	"log/slog"
+	"net/netip"
+	"sync"
+	"time"
+
+	"go4.org/netipx"
+)
+
+// Format is how a list's file is written.
+type Format int
+
+const (
+	// FormatLines has one address or network per line, such as 192.0.2.1 or
+	// 192.0.2.0/24. Lines that start with # are comments.
+	FormatLines Format = iota
+
+	// FormatSpamhausJSON has one JSON object per line, with a network in its
+	// cidr field. The last line holds the file's metadata and no network.
+	FormatSpamhausJSON
+
+	// FormatDShield has one network per line, as tab-separated fields that
+	// start with the network's first and last address. Lines that start
+	// with # are comments.
+	FormatDShield
+)
+
+// List is one public blocklist.
+type List struct {
+	// Name names the list in lookup answers, and names its copy on disk.
+	Name string
+
+	// Description says in one line what being on the list means.
+	Description string
+
+	// URL is where the publisher serves the whole list.
+	URL string
+
+	// Interval is how long a copy is used before the list is downloaded
+	// again.
+	Interval time.Duration
+
+	// Format is how the list's file is written.
+	Format Format
+}
+
+// Listing says that a list contains an address.
+type Listing struct {
+	Name        string `json:"name"`
+	Description string `json:"description"`
+
+	// Updated is when routewatch last downloaded the list, or last heard from
+	// its publisher that its copy was still current.
+	Updated time.Time `json:"updated"`
+}
+
+// listCopy is a list's last good copy in memory.
+type listCopy struct {
+	addresses *netipx.IPSet
+	updated   time.Time
+
+	// validators identify the version of the list this copy holds.
+	validators validators
+}
+
+// Params is what New needs.
+type Params struct {
+	// Dir is the directory that holds the copies on disk.
+	Dir string
+
+	// Lists are the lists to download and look addresses up in.
+	Lists []List
+
+	Logger *slog.Logger
+}
+
+// Blocklists keeps a copy of each list and answers which lists contain an
+// address. Start loads the copies from disk and starts downloading the lists.
+type Blocklists struct {
+	dir    string
+	lists  []List
+	logger *slog.Logger
+
+	// copies holds each list's last good copy, by list name. A list with no
+	// good copy yet has none.
+	mu     sync.RWMutex
+	copies map[string]listCopy
+
+	// stop ends the goroutine Start runs, and done is closed when it returns.
+	stop context.CancelFunc
+	done chan struct{}
+}
+
+// New returns a Blocklists with no copies. Start loads them.
+func New(params Params) *Blocklists {
+	return &Blocklists{
+		dir:    params.Dir,
+		lists:  params.Lists,
+		logger: params.Logger.With("component", "blocklists"),
+		copies: make(map[string]listCopy),
+		done:   make(chan struct{}),
+	}
+}
+
+// Lookup returns a Listing for each list whose copy contains addr, in the
+// order of the lists given to New. It makes no outside request. When no list
+// contains addr, it returns an empty list, not nil.
+func (b *Blocklists) Lookup(addr netip.Addr) []Listing {
+	b.mu.RLock()
+	defer b.mu.RUnlock()
+
+	listings := []Listing{}
+
+	for _, list := range b.lists {
+		c, ok := b.copies[list.Name]
+		if ok && c.addresses.Contains(addr.Unmap()) {
+			listings = append(listings, Listing{
+				Name:        list.Name,
+				Description: list.Description,
+				Updated:     c.updated,
+			})
+		}
+	}
+
+	return listings
+}
+
+// ShortestInterval returns the shortest Interval of the lists. That is the
+// soonest a copy can change, and so how long a lookup answer stays current.
+func (b *Blocklists) ShortestInterval() time.Duration {
+	var shortest time.Duration
+
+	for i, list := range b.lists {
+		if i == 0 || list.Interval < shortest {
+			shortest = list.Interval
+		}
+	}
+
+	return shortest
+}
+
+// updated returns when the named list's copy was last downloaded or found
+// current, or the zero time when the list has no copy.
+func (b *Blocklists) updated(name string) time.Time {
+	b.mu.RLock()
+	defer b.mu.RUnlock()
+
+	return b.copies[name].updated
+}
+
+// copyValidators returns the validators of the named list's copy. They are
+// empty when the list has no copy.
+func (b *Blocklists) copyValidators(name string) validators {
+	b.mu.RLock()
+	defer b.mu.RUnlock()
+
+	return b.copies[name].validators
+}
+
+// setCopy makes c the named list's copy.
+func (b *Blocklists) setCopy(name string, c listCopy) {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	b.copies[name] = c
+}
+
+// setUpdated records that the named list's copy was found current at t.
+func (b *Blocklists) setUpdated(name string, t time.Time) {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	c, ok := b.copies[name]
+	if ok {
+		c.updated = t
+		b.copies[name] = c
+	}
+}

internal/blocklist/blocklist_internal_test.go

@@ -0,0 +1,494 @@
+package blocklist
+
+import (
+	"context"
+	"errors"
+	"io"
+	"io/fs"
+	"log/slog"
+	"net/http"
+	"net/http/httptest"
+	"net/netip"
+	"os"
+	"path/filepath"
+	"strings"
+	"sync"
+	"testing"
+	"time"
+)
+
+// The test list's name and description, the address its good version lists,
+// and the address a newer version lists instead.
+const (
+	testListName        = "test_list"
+	testListDescription = "On the test list"
+	listedAddr          = "192.0.2.1"
+	goodListBody        = "# A test list\n" + listedAddr + "\n"
+	newListedAddr       = "192.0.2.2"
+	newListBody         = "# A test list\n" + newListedAddr + "\n"
+)
+
+// testETag is the ETag a publisher sends with a version of the test list.
+const testETag = `"v1"`
+
+// maxWait is how long a test waits for the download goroutine.
+const maxWait = 10 * time.Second
+
+// publisher stands in for a list's publisher. It answers every request with
+// its status and body, and keeps the headers of each request. A 200 answer
+// carries lastModified and etag, when set, as Last-Modified and ETag. As
+// from a real publisher, a request whose If-None-Match or If-Modified-Since
+// matches them is answered "not modified".
+type publisher struct {
+	mu           sync.Mutex
+	status       int
+	body         string
+	lastModified time.Time
+	etag         string
+	headers      []http.Header
+}
+
+// newPublisher starts a publisher that answers 200 with body, and returns it
+// and its URL.
+func newPublisher(t *testing.T, body string) (*publisher, string) {
+	t.Helper()
+
+	p := &publisher{status: http.StatusOK, body: body}
+	srv := httptest.NewServer(p)
+	t.Cleanup(srv.Close)
+
+	return p, srv.URL
+}
+
+func (p *publisher) ServeHTTP(w http.ResponseWriter, r *http.Request) {
+	p.mu.Lock()
+	defer p.mu.Unlock()
+
+	p.headers = append(p.headers, r.Header.Clone())
+
+	if p.status != http.StatusOK {
+		w.WriteHeader(p.status)
+		_, _ = io.WriteString(w, p.body)
+
+		return
+	}
+
+	if p.etag != "" {
+		w.Header().Set(etagHeader, p.etag)
+	}
+
+	// ServeContent sends Last-Modified unless lastModified is zero, and
+	// answers "not modified" when the request's conditions match.
+	http.ServeContent(w, r, testListName, p.lastModified,
+		strings.NewReader(p.body))
+}
+
+// answer makes the publisher answer later requests with status and body.
+func (p *publisher) answer(status int, body string) {
+	p.mu.Lock()
+	defer p.mu.Unlock()
+
+	p.status, p.body = status, body
+}
+
+// publish makes the publisher answer later requests with a version of the
+// list: body, with lastModified and etag. A zero lastModified or an empty
+// etag is not sent.
+func (p *publisher) publish(body string, lastModified time.Time, etag string) {
+	p.mu.Lock()
+	defer p.mu.Unlock()
+
+	p.status, p.body = http.StatusOK, body
+	p.lastModified, p.etag = lastModified, etag
+}
+
+// requests returns the headers of the requests the publisher has had.
+func (p *publisher) requests() []http.Header {
+	p.mu.Lock()
+	defer p.mu.Unlock()
+
+	return append([]http.Header(nil), p.headers...)
+}
+
+// testList returns a list served at url, written in format.
+func testList(url string, format Format) List {
+	return List{
+		Name:        testListName,
+		Description: testListDescription,
+		URL:         url,
+		Interval:    time.Hour,
+		Format:      format,
+	}
+}
+
+func newTestBlocklists(dir string, list List) *Blocklists {
+	return New(Params{
+		Dir:    dir,
+		Lists:  []List{list},
+		Logger: slog.New(slog.DiscardHandler),
+	})
+}
+
+// lookup looks addr up in b.
+func lookup(t *testing.T, b *Blocklists, addr string) []Listing {
+	t.Helper()
+
+	return b.Lookup(netip.MustParseAddr(addr))
+}
+
+// TestLookupFindsListedAddresses downloads a list in each format and looks
+// up addresses on it, inside its networks, and off it, in IPv4 and IPv6.
+// A listed address has one listing that names the list; any other has an
+// empty list of listings.
+func TestLookupFindsListedAddresses(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name     string
+		format   Format
+		body     string
+		listed   []string
+		unlisted []string
+	}{
+		{
+			name:     "single addresses",
+			format:   FormatLines,
+			body:     "# comment\n192.0.2.1\n\n2001:db8::1\n",
+			listed:   []string{"192.0.2.1", "::ffff:192.0.2.1", "2001:db8::1"},
+			unlisted: []string{"192.0.2.2", "2001:db8::2"},
+		},
+		{
+			name:     "networks",
+			format:   FormatLines,
+			body:     "# comment\n198.51.100.0/24\n2001:db8:1::/48\n",
+			listed:   []string{"198.51.100.0", "198.51.100.255", "2001:db8:1::5"},
+			unlisted: []string{"198.51.101.0", "2001:db8:2::1"},
+		},
+		{
+			name:   "Spamhaus networks",
+			format: FormatSpamhausJSON,
+			body: `{"cidr":"203.0.113.0/25","sblid":"SBL1","rir":"arin"}` + "\n" +
+				`{"cidr":"2001:db8:3::/48","sblid":"SBL2","rir":"ripencc"}` + "\n" +
+				`{"type":"metadata","timestamp":1793074442,"records":2}` + "\n",
+			listed:   []string{"203.0.113.1", "2001:db8:3::1"},
+			unlisted: []string{"203.0.113.200", "2001:db8:4::1"},
+		},
+		{
+			name:   "DShield networks",
+			format: FormatDShield,
+			body: "#   DShield.org Recommended Block List\n" +
+				"192.0.2.0\t192.0.2.255\t24\t327\tEXAMPLE\tUS\tNone\n",
+			listed:   []string{"192.0.2.0", "192.0.2.77"},
+			unlisted: []string{"192.0.3.1", "2001:db8::1"},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			_, url := newPublisher(t, tt.body)
+			b := newTestBlocklists(t.TempDir(), testList(url, tt.format))
+
+			err := b.download(t.Context(), b.lists[0])
+			if err != nil {
+				t.Fatalf("download: %v", err)
+			}
+
+			for _, addr := range tt.listed {
+				listings := lookup(t, b, addr)
+				if len(listings) != 1 || listings[0].Name != testListName ||
+					listings[0].Description != testListDescription ||
+					listings[0].Updated.IsZero() {
+					t.Errorf("%s: listings %+v, want one for %s", addr,
+						listings, testListName)
+				}
+			}
+
+			for _, addr := range tt.unlisted {
+				listings := lookup(t, b, addr)
+				if listings == nil || len(listings) != 0 {
+					t.Errorf("%s: listings %#v, want an empty list", addr, listings)
+				}
+			}
+		})
+	}
+}
+
+// TestFailedDownloadKeepsThePreviousCopy downloads a good version of a list,
+// then one that fails. The good copy stays in memory and on disk, and no
+// part of the failed download is left beside it.
+func TestFailedDownloadKeepsThePreviousCopy(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name   string
+		status int
+		body   string
+	}{
+		{"server error", http.StatusInternalServerError, "unavailable"},
+		{"not a list", http.StatusOK, "<!DOCTYPE html>\n<html></html>\n"},
+		{
+			"over the size limit", http.StatusOK,
+			strings.Repeat(listedAddr+"\n", maxDownloadSize/len(listedAddr)),
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			dir := t.TempDir()
+			p, url := newPublisher(t, goodListBody)
+			b := newTestBlocklists(dir, testList(url, FormatLines))
+
+			err := b.download(t.Context(), b.lists[0])
+			if err != nil {
+				t.Fatalf("first download: %v", err)
+			}
+
+			before := lookup(t, b, listedAddr)
+
+			p.answer(tt.status, tt.body)
+
+			err = b.download(t.Context(), b.lists[0])
+			if err == nil {
+				t.Fatal("second download returned no error")
+			}
+
+			after := lookup(t, b, listedAddr)
+			if len(after) != 1 || len(before) != 1 ||
+				!after[0].Updated.Equal(before[0].Updated) {
+				t.Errorf("listings %+v after the failure, want %+v", after, before)
+			}
+
+			copies := os.DirFS(dir)
+
+			onDisk, err := fs.ReadFile(copies, testListName)
+			if err != nil || string(onDisk) != goodListBody {
+				t.Errorf("copy on disk %q (%v), want %q", onDisk, err, goodListBody)
+			}
+
+			_, err = fs.Stat(copies, testListName+downloadSuffix)
+			if !errors.Is(err, fs.ErrNotExist) {
+				t.Errorf("the failed download's file is still there (%v)", err)
+			}
+		})
+	}
+}
+
+// TestRestartLoadsCopiesFromDisk downloads a list, then starts a new
+// Blocklists on the same directory, as a restart does. The new one answers
+// from the copy on disk at once, and does not download the list again while
+// the copy is younger than the list's interval.
+func TestRestartLoadsCopiesFromDisk(t *testing.T) {
+	t.Parallel()
+
+	dir := t.TempDir()
+	p, url := newPublisher(t, goodListBody)
+	list := testList(url, FormatLines)
+
+	first := newTestBlocklists(dir, list)
+
+	err := first.download(t.Context(), list)
+	if err != nil {
+		t.Fatalf("download: %v", err)
+	}
+
+	downloaded := lookup(t, first, listedAddr)[0].Updated
+
+	second := newTestBlocklists(dir, list)
+	second.Start(t.Context())
+	second.Stop(t.Context())
+
+	listings := lookup(t, second, listedAddr)
+	if len(listings) != 1 ||
+		listings[0].Updated.Sub(downloaded).Abs() > time.Second {
+		t.Errorf("after the restart, listings %+v, want one updated at %s",
+			listings, downloaded)
+	}
+
+	if got := len(p.requests()); got != 1 {
+		t.Errorf("the publisher had %d requests, want 1", got)
+	}
+}
+
+// TestStartRefreshesAnOldCopy starts on a copy older than its list's
+// interval. The list is downloaded at once, with routewatch's User-Agent and
+// the Last-Modified and ETag the publisher sent with the copy, kept on disk.
+// The publisher answers that the copy is still current, so the copy is kept
+// and marked as updated now.
+func TestStartRefreshesAnOldCopy(t *testing.T) {
+	t.Parallel()
+
+	dir := t.TempDir()
+	p, url := newPublisher(t, goodListBody)
+	list := testList(url, FormatLines)
+
+	lastModified := time.Now().Add(-3 * list.Interval).UTC().Truncate(time.Second)
+	p.publish(goodListBody, lastModified, testETag)
+
+	err := newTestBlocklists(dir, list).download(t.Context(), list)
+	if err != nil {
+		t.Fatalf("download: %v", err)
+	}
+
+	old := time.Now().Add(-2 * list.Interval).UTC().Truncate(time.Second)
+
+	err = os.Chtimes(filepath.Join(dir, testListName), old, old)
+	if err != nil {
+		t.Fatalf("Chtimes: %v", err)
+	}
+
+	b := newTestBlocklists(dir, list)
+	b.Start(t.Context())
+
+	deadline := time.Now().Add(maxWait)
+	for !lookup(t, b, listedAddr)[0].Updated.After(old) {
+		if time.Now().After(deadline) {
+			t.Fatalf("the copy was not marked updated within %s", maxWait)
+		}
+
+		time.Sleep(time.Millisecond)
+	}
+
+	b.Stop(t.Context())
+
+	requests := p.requests()
+	last := requests[len(requests)-1]
+
+	want := lastModified.Format(http.TimeFormat)
+	if got := last.Get(ifModifiedSinceHeader); got != want {
+		t.Errorf("If-Modified-Since %q, want %q", got, want)
+	}
+
+	if got := last.Get(ifNoneMatchHeader); got != testETag {
+		t.Errorf("If-None-Match %q, want %q", got, testETag)
+	}
+
+	if got := last.Get(userAgentHeader); got != userAgent {
+		t.Errorf("User-Agent %q, want %q", got, userAgent)
+	}
+}
+
+// TestNewVersionOlderThanTheLastDownloadReplacesTheCopy downloads a list.
+// The publisher then puts out a new version whose Last-Modified is earlier
+// than that download, as when a cache in front of the publisher passes the
+// version on late. The next download sends the Last-Modified of the copy,
+// not the time of the download, so it gets the new version.
+func TestNewVersionOlderThanTheLastDownloadReplacesTheCopy(t *testing.T) {
+	t.Parallel()
+
+	p, url := newPublisher(t, goodListBody)
+	list := testList(url, FormatLines)
+	b := newTestBlocklists(t.TempDir(), list)
+
+	firstVersion := time.Now().Add(-2 * list.Interval).UTC().Truncate(time.Second)
+	p.publish(goodListBody, firstVersion, "")
+
+	err := b.download(t.Context(), list)
+	if err != nil {
+		t.Fatalf("first download: %v", err)
+	}
+
+	p.publish(newListBody, firstVersion.Add(list.Interval), "")
+
+	err = b.download(t.Context(), list)
+	if err != nil {
+		t.Fatalf("second download: %v", err)
+	}
+
+	if listings := lookup(t, b, newListedAddr); len(listings) != 1 {
+		t.Errorf("%s: listings %+v, want one from the new version",
+			newListedAddr, listings)
+	}
+
+	want := firstVersion.Format(http.TimeFormat)
+	if got := p.requests()[1].Get(ifModifiedSinceHeader); got != want {
+		t.Errorf("If-Modified-Since %q, want %q", got, want)
+	}
+}
+
+// TestDownloadWithoutLastModifiedOrETagSendsNoCondition downloads a list
+// twice from a publisher that sends neither Last-Modified nor ETag. Neither
+// request carries If-Modified-Since or If-None-Match, so the second download
+// gets the whole list again.
+func TestDownloadWithoutLastModifiedOrETagSendsNoCondition(t *testing.T) {
+	t.Parallel()
+
+	p, url := newPublisher(t, goodListBody)
+	list := testList(url, FormatLines)
+	b := newTestBlocklists(t.TempDir(), list)
+
+	err := b.download(t.Context(), list)
+	if err != nil {
+		t.Fatalf("first download: %v", err)
+	}
+
+	err = b.download(t.Context(), list)
+	if err != nil {
+		t.Fatalf("second download: %v", err)
+	}
+
+	for i, request := range p.requests() {
+		if request.Get(ifModifiedSinceHeader) != "" ||
+			request.Get(ifNoneMatchHeader) != "" {
+			t.Errorf("request %d has a condition: %v", i, request)
+		}
+	}
+}
+
+// TestStopCancelsADownloadInProgress stops while the publisher has not
+// answered. Stop cancels the download and returns once the download
+// goroutine has stopped.
+func TestStopCancelsADownloadInProgress(t *testing.T) {
+	t.Parallel()
+
+	arrived := make(chan struct{})
+
+	var once sync.Once
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(_ http.ResponseWriter, r *http.Request) {
+			once.Do(func() { close(arrived) })
+			<-r.Context().Done()
+		}))
+	t.Cleanup(srv.Close)
+
+	b := newTestBlocklists(t.TempDir(), testList(srv.URL, FormatLines))
+	b.Start(t.Context())
+
+	select {
+	case <-arrived:
+	case <-time.After(maxWait):
+		t.Fatalf("no download began within %s", maxWait)
+	}
+
+	ctx, cancel := context.WithTimeout(t.Context(), maxWait)
+	defer cancel()
+
+	b.Stop(ctx)
+
+	select {
+	case <-b.done:
+	default:
+		t.Error("Stop returned before the download goroutine stopped")
+	}
+}
+
+// TestPublicListsAreComplete checks that every public list has what a
+// download and a lookup answer need, and a unique name.
+func TestPublicListsAreComplete(t *testing.T) {
+	t.Parallel()
+
+	names := make(map[string]bool)
+
+	for _, list := range Lists() {
+		if list.Name == "" || list.Description == "" || list.URL == "" ||
+			list.Interval <= 0 || names[list.Name] {
+			t.Errorf("list %+v is incomplete or has a name used before", list)
+		}
+
+		names[list.Name] = true
+	}
+}

internal/blocklist/download.go

@@ -0,0 +1,379 @@
+package blocklist
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"io/fs"
+	"net/http"
+	"os"
+	"slices"
+	"time"
+
+	"go4.org/netipx"
+)
+
+// Limits of one download.
+const (
+	// downloadTimeout is how long one download may take.
+	downloadTimeout = time.Minute
+
+	// maxDownloadSize is the size of the largest list file accepted, in
+	// bytes. The largest list, FireHOL abusers, is about 2.5 MiB.
+	maxDownloadSize = 16 << 20
+)
+
+// The request headers a download sets.
+const (
+	userAgentHeader       = "User-Agent"
+	ifModifiedSinceHeader = "If-Modified-Since"
+	ifNoneMatchHeader     = "If-None-Match"
+
+	// userAgent names routewatch to the publishers.
+	userAgent = "routewatch (+https://git.eeqj.de/sneak/routewatch)"
+)
+
+// The response headers that identify the version of a list a publisher sent.
+const (
+	lastModifiedHeader = "Last-Modified"
+	etagHeader         = "ETag"
+)
+
+const (
+	// downloadSuffix ends the name of the file a download is written to,
+	// beside the list's copy, until it has parsed.
+	downloadSuffix = ".download"
+
+	// validatorsSuffix ends the name of the file beside a list's copy that
+	// keeps the copy's validators.
+	validatorsSuffix = ".validators"
+
+	// dirPermissions are those of the directory that holds the copies.
+	dirPermissions = 0o750
+
+	// validatorsPermissions are those of the file that keeps a copy's
+	// validators.
+	validatorsPermissions = 0o600
+)
+
+// validators are the Last-Modified and ETag headers a publisher sent with a
+// version of a list. A download sends them back, as If-Modified-Since and
+// If-None-Match, so that the publisher sends the list only when it has
+// changed. They are kept as the publisher wrote them.
+//
+//nolint:tagliatelle // snake_case keys, as the Go style guide asks
+type validators struct {
+	LastModified string `json:"last_modified,omitempty"`
+	ETag         string `json:"etag,omitempty"`
+}
+
+var (
+	errUnexpectedStatus = errors.New("unexpected HTTP status")
+	errTooLarge         = errors.New("list is over the size limit")
+)
+
+// Start loads each list's copy from disk. It then starts the goroutine that
+// downloads each list once its copy is older than its Interval, and a list
+// with no copy at once. The goroutine runs until Stop: the end of ctx does
+// not stop it.
+func (b *Blocklists) Start(ctx context.Context) {
+	err := os.MkdirAll(b.dir, dirPermissions)
+	if err != nil {
+		b.logger.Error("Failed to create the blocklists directory",
+			"dir", b.dir, "error", err)
+	}
+
+	nextDownload := make([]time.Time, len(b.lists))
+
+	for i, list := range b.lists {
+		loadErr := b.loadCopy(list)
+		if loadErr != nil && !errors.Is(loadErr, fs.ErrNotExist) {
+			b.logger.Error("Failed to load a blocklist copy",
+				"list", list.Name, "error", loadErr)
+		}
+
+		// A list with no copy has the zero time, so it is due at once.
+		nextDownload[i] = b.updated(list.Name).Add(list.Interval)
+	}
+
+	ctx, b.stop = context.WithCancel(context.WithoutCancel(ctx))
+
+	go func() {
+		b.run(ctx, nextDownload)
+	}()
+
+	b.logger.Info("Blocklist downloads started", "dir", b.dir)
+}
+
+// Stop stops the downloads, cancelling one in progress, and waits until the
+// goroutine has returned or ctx ends. The copies in memory stay, so Lookup
+// still answers.
+func (b *Blocklists) Stop(ctx context.Context) {
+	b.stop()
+
+	select {
+	case <-b.done:
+		b.logger.Info("Blocklist downloads stopped")
+	case <-ctx.Done():
+		b.logger.Warn("Stop deadline passed before the blocklist downloads " +
+			"stopped")
+	}
+}
+
+// run downloads each list when nextDownload says, until ctx ends. After each
+// try, whether it worked or not, the list's next download is an Interval
+// later.
+func (b *Blocklists) run(ctx context.Context, nextDownload []time.Time) {
+	defer close(b.done)
+
+	if len(nextDownload) == 0 {
+		return
+	}
+
+	for {
+		soonest := slices.MinFunc(nextDownload, time.Time.Compare)
+		timer := time.NewTimer(time.Until(soonest))
+
+		select {
+		case <-ctx.Done():
+			timer.Stop()
+
+			return
+		case <-timer.C:
+		}
+
+		for i, list := range b.lists {
+			if time.Now().Before(nextDownload[i]) {
+				continue
+			}
+
+			err := b.download(ctx, list)
+			if ctx.Err() != nil {
+				return
+			}
+
+			if err != nil {
+				b.logger.Error("Blocklist download failed; keeping the "+
+					"previous copy", "list", list.Name, "error", err)
+			}
+
+			nextDownload[i] = time.Now().Add(list.Interval)
+		}
+	}
+}
+
+// download asks the publisher for list, sending the validators of the list's
+// copy when it has them. A new version that parses becomes the list's copy,
+// on disk and in memory. An answer that the copy is still current marks the
+// copy as updated now. Anything else leaves the copy as it is and returns an
+// error.
+func (b *Blocklists) download(ctx context.Context, list List) error {
+	ctx, cancel := context.WithTimeout(ctx, downloadTimeout)
+	defer cancel()
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, list.URL, nil)
+	if err != nil {
+		return err
+	}
+
+	req.Header.Set(userAgentHeader, userAgent)
+
+	held := b.copyValidators(list.Name)
+	if held.LastModified != "" {
+		req.Header.Set(ifModifiedSinceHeader, held.LastModified)
+	}
+
+	if held.ETag != "" {
+		req.Header.Set(ifNoneMatchHeader, held.ETag)
+	}
+
+	resp, err := http.DefaultClient.Do(req)
+	if err != nil {
+		return err
+	}
+
+	defer func() { _ = resp.Body.Close() }()
+
+	switch resp.StatusCode {
+	case http.StatusOK:
+		return b.replaceCopy(list, resp.Body, validators{
+			LastModified: resp.Header.Get(lastModifiedHeader),
+			ETag:         resp.Header.Get(etagHeader),
+		})
+	case http.StatusNotModified:
+		return b.markCurrent(list)
+	default:
+		return fmt.Errorf("%w: %s", errUnexpectedStatus, resp.Status)
+	}
+}
+
+// replaceCopy writes a new version of list, read from body, beside the list's
+// copy. Only when it parses does it replace the copy, on disk and in memory.
+// sent are the validators the publisher sent with the new version.
+func (b *Blocklists) replaceCopy(list List, body io.Reader, sent validators) error {
+	root, err := os.OpenRoot(b.dir)
+	if err != nil {
+		return err
+	}
+
+	defer func() { _ = root.Close() }()
+
+	downloadName := list.Name + downloadSuffix
+
+	addresses, err := writeAndParse(root, downloadName, body, list.Format)
+	if err != nil {
+		_ = root.Remove(downloadName)
+
+		return err
+	}
+
+	// The old copy's validators are removed before the copy is replaced, so
+	// that they are never kept with the new copy.
+	err = root.Remove(list.Name + validatorsSuffix)
+	if err != nil && !errors.Is(err, fs.ErrNotExist) {
+		return err
+	}
+
+	err = root.Rename(downloadName, list.Name)
+	if err != nil {
+		return err
+	}
+
+	b.setCopy(list.Name, listCopy{
+		addresses: addresses, updated: time.Now().UTC(), validators: sent,
+	})
+	b.logger.Info("Downloaded a new version of a blocklist", "list", list.Name)
+
+	// Without this file, the first download after a restart asks for the
+	// whole list.
+	err = writeValidators(root, list.Name, sent)
+	if err != nil {
+		b.logger.Error("Failed to keep the validators of a blocklist copy",
+			"list", list.Name, "error", err)
+	}
+
+	return nil
+}
+
+// writeAndParse writes body to the file name in root, up to maxDownloadSize
+// bytes, and parses the file as format.
+func writeAndParse(
+	root *os.Root, name string, body io.Reader, format Format,
+) (*netipx.IPSet, error) {
+	f, err := root.Create(name)
+	if err != nil {
+		return nil, err
+	}
+
+	defer func() { _ = f.Close() }()
+
+	written, err := io.Copy(f, io.LimitReader(body, maxDownloadSize+1))
+	if err != nil {
+		return nil, err
+	}
+
+	if written > maxDownloadSize {
+		return nil, errTooLarge
+	}
+
+	_, err = f.Seek(0, io.SeekStart)
+	if err != nil {
+		return nil, err
+	}
+
+	return parse(f, format)
+}
+
+// markCurrent records that the publisher says the list's copy is still
+// current: the copy, on disk and in memory, is marked as updated now.
+func (b *Blocklists) markCurrent(list List) error {
+	root, err := os.OpenRoot(b.dir)
+	if err != nil {
+		return err
+	}
+
+	defer func() { _ = root.Close() }()
+
+	now := time.Now().UTC()
+
+	err = root.Chtimes(list.Name, now, now)
+	if err != nil {
+		return err
+	}
+
+	b.setUpdated(list.Name, now)
+	b.logger.Debug("Blocklist copy is still current", "list", list.Name)
+
+	return nil
+}
+
+// loadCopy makes the list's copy on disk, with the validators kept beside it,
+// its copy in memory. The file's modification time is when the copy was
+// updated.
+func (b *Blocklists) loadCopy(list List) error {
+	root, err := os.OpenRoot(b.dir)
+	if err != nil {
+		return err
+	}
+
+	defer func() { _ = root.Close() }()
+
+	f, err := root.Open(list.Name)
+	if err != nil {
+		return err
+	}
+
+	defer func() { _ = f.Close() }()
+
+	info, err := f.Stat()
+	if err != nil {
+		return err
+	}
+
+	addresses, err := parse(f, list.Format)
+	if err != nil {
+		return err
+	}
+
+	held, err := readValidators(root, list.Name)
+	if err != nil {
+		return err
+	}
+
+	b.setCopy(list.Name, listCopy{
+		addresses: addresses, updated: info.ModTime().UTC(), validators: held,
+	})
+
+	return nil
+}
+
+// writeValidators keeps held in the file beside the named list's copy.
+func writeValidators(root *os.Root, name string, held validators) error {
+	data, err := json.Marshal(held)
+	if err != nil {
+		return err
+	}
+
+	return root.WriteFile(name+validatorsSuffix, data, validatorsPermissions)
+}
+
+// readValidators returns the validators kept beside the named list's copy.
+// They are empty when there is no such file.
+func readValidators(root *os.Root, name string) (validators, error) {
+	var held validators
+
+	data, err := root.ReadFile(name + validatorsSuffix)
+	if errors.Is(err, fs.ErrNotExist) {
+		return held, nil
+	}
+
+	if err != nil {
+		return held, err
+	}
+
+	err = json.Unmarshal(data, &held)
+
+	return held, err
+}

internal/blocklist/lists.go

@@ -0,0 +1,93 @@
+package blocklist
+
+import "time"
+
+// How often each list is downloaded. No list is downloaded more often than
+// its publisher updates it or allows.
+const (
+	// DShield asks that its feeds be downloaded at most once an hour:
+	// https://isc.sans.edu/feeds_doc.html
+	dshieldInterval = time.Hour
+
+	// abuse.ch makes the list every 5 minutes and recommends a download at
+	// least every 15: https://feodotracker.abuse.ch/blocklist/
+	feodoTrackerInterval = 15 * time.Minute
+
+	// FireHOL's page for each list gives how often the list changes on
+	// average: about every 5 hours for abusers and every 3 hours for level
+	// 3. FireHOL sets no download limit.
+	// https://iplists.firehol.org/?ipset=firehol_abusers_30d
+	// https://iplists.firehol.org/?ipset=firehol_level3
+	fireHOLAbusersInterval = 5 * time.Hour
+	fireHOLLevel3Interval  = 3 * time.Hour
+
+	// Spamhaus asks that DROP be downloaded at most once a day:
+	// https://www.spamhaus.org/blocklists/do-not-route-or-peer/
+	spamhausDROPInterval = 24 * time.Hour
+
+	// Tor publishes a new exit list every hour and sets no download limit:
+	// https://collector.torproject.org/recent/exit-lists/
+	torExitsInterval = time.Hour
+)
+
+// Lists returns the public blocklists routewatch downloads. Each can be
+// downloaded whole without an account or a key. Spamhaus merged its EDROP
+// list into DROP, so DROP covers both.
+func Lists() []List {
+	return []List{
+		{
+			Name: "dshield",
+			Description: "In one of the 20 networks that DShield saw scan " +
+				"the most targets in the last three days",
+			URL:      "https://feeds.dshield.org/block.txt",
+			Interval: dshieldInterval,
+			Format:   FormatDShield,
+		},
+		{
+			Name:        "feodo_tracker",
+			Description: "A botnet command-and-control server",
+			URL:         "https://feodotracker.abuse.ch/downloads/ipblocklist.txt",
+			Interval:    feodoTrackerInterval,
+			Format:      FormatLines,
+		},
+		{
+			Name: "firehol_abusers_30d",
+			Description: "Reported for comment, form or forum spam in the " +
+				"last 30 days",
+			URL:      "https://iplists.firehol.org/files/firehol_abusers_30d.netset",
+			Interval: fireHOLAbusersInterval,
+			Format:   FormatLines,
+		},
+		{
+			Name: "firehol_level3",
+			Description: "Reported for attacks, spyware or viruses in the " +
+				"last 30 days",
+			URL:      "https://iplists.firehol.org/files/firehol_level3.netset",
+			Interval: fireHOLLevel3Interval,
+			Format:   FormatLines,
+		},
+		{
+			Name: "spamhaus_drop",
+			Description: "In an IPv4 network that Spamhaus says spammers " +
+				"or cyber criminals have hijacked or lease",
+			URL:      "https://www.spamhaus.org/drop/drop_v4.json",
+			Interval: spamhausDROPInterval,
+			Format:   FormatSpamhausJSON,
+		},
+		{
+			Name: "spamhaus_dropv6",
+			Description: "In an IPv6 network that Spamhaus says spammers " +
+				"or cyber criminals have hijacked or lease",
+			URL:      "https://www.spamhaus.org/drop/drop_v6.json",
+			Interval: spamhausDROPInterval,
+			Format:   FormatSpamhausJSON,
+		},
+		{
+			Name:        "tor_exits",
+			Description: "A Tor exit node, where traffic from the Tor network leaves",
+			URL:         "https://check.torproject.org/torbulkexitlist",
+			Interval:    torExitsInterval,
+			Format:      FormatLines,
+		},
+	}
+}

internal/blocklist/parse.go

@@ -0,0 +1,158 @@
+package blocklist
+
+import (
+	"bufio"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/netip"
+	"strings"
+
+	"go4.org/netipx"
+)
+
+const (
+	// commentPrefix starts a comment line in a list file.
+	commentPrefix = "#"
+
+	// networkSeparator separates the address and the mask length of a
+	// network, as in 192.0.2.0/24.
+	networkSeparator = "/"
+
+	// dshieldFieldSeparator separates the fields of a DShield line.
+	dshieldFieldSeparator = "\t"
+
+	// spamhausMetadataType is the type of the line that ends a Spamhaus
+	// file. That line holds the file's metadata and no network.
+	spamhausMetadataType = "metadata"
+)
+
+var (
+	errUnknownFormat = errors.New("unknown list format")
+	errNoNetwork     = errors.New("no network on the line")
+	errBadRange      = errors.New("first and last address make no range")
+)
+
+// parse reads a list file written in format and returns the addresses it
+// holds. A line it cannot read fails the whole file, so a file that is not
+// the list, such as an error page, never replaces a good copy.
+func parse(r io.Reader, format Format) (*netipx.IPSet, error) {
+	var builder netipx.IPSetBuilder
+
+	scanner := bufio.NewScanner(r)
+
+	for lineNumber := 1; scanner.Scan(); lineNumber++ {
+		line := strings.TrimSpace(scanner.Text())
+		if line == "" || strings.HasPrefix(line, commentPrefix) {
+			continue
+		}
+
+		err := addLine(&builder, line, format)
+		if err != nil {
+			return nil, fmt.Errorf("line %d: %w", lineNumber, err)
+		}
+	}
+
+	err := scanner.Err()
+	if err != nil {
+		return nil, err
+	}
+
+	return builder.IPSet()
+}
+
+// addLine adds the address or network on one line of a list file to builder.
+func addLine(builder *netipx.IPSetBuilder, line string, format Format) error {
+	switch format {
+	case FormatLines:
+		return addAddressOrNetwork(builder, line)
+	case FormatSpamhausJSON:
+		return addSpamhausNetwork(builder, line)
+	case FormatDShield:
+		return addDShieldNetwork(builder, line)
+	default:
+		return errUnknownFormat
+	}
+}
+
+// addAddressOrNetwork adds an address, such as 192.0.2.1, or a network, such
+// as 192.0.2.0/24, to builder.
+func addAddressOrNetwork(builder *netipx.IPSetBuilder, text string) error {
+	if strings.Contains(text, networkSeparator) {
+		network, err := netip.ParsePrefix(text)
+		if err != nil {
+			return err
+		}
+
+		builder.AddPrefix(network)
+
+		return nil
+	}
+
+	addr, err := netip.ParseAddr(text)
+	if err != nil {
+		return err
+	}
+
+	builder.Add(addr)
+
+	return nil
+}
+
+// addSpamhausNetwork adds the network on a line of a Spamhaus file to
+// builder. The metadata line adds nothing.
+func addSpamhausNetwork(builder *netipx.IPSetBuilder, line string) error {
+	var entry struct {
+		CIDR string `json:"cidr"`
+		Type string `json:"type"`
+	}
+
+	err := json.Unmarshal([]byte(line), &entry)
+	if err != nil {
+		return err
+	}
+
+	if entry.Type == spamhausMetadataType {
+		return nil
+	}
+
+	network, err := netip.ParsePrefix(entry.CIDR)
+	if err != nil {
+		return err
+	}
+
+	builder.AddPrefix(network)
+
+	return nil
+}
+
+// addDShieldNetwork adds the network on a line of a DShield file to builder.
+// The line's first two fields are the network's first and last address.
+func addDShieldNetwork(builder *netipx.IPSetBuilder, line string) error {
+	firstField, rest, found := strings.Cut(line, dshieldFieldSeparator)
+	if !found {
+		return errNoNetwork
+	}
+
+	lastField, _, _ := strings.Cut(rest, dshieldFieldSeparator)
+
+	first, err := netip.ParseAddr(firstField)
+	if err != nil {
+		return err
+	}
+
+	last, err := netip.ParseAddr(lastField)
+	if err != nil {
+		return err
+	}
+
+	network := netipx.IPRangeFrom(first, last)
+	if !network.IsValid() {
+		return errBadRange
+	}
+
+	builder.AddRange(network)
+
+	return nil
+}

internal/config/config.go

@@ -2,11 +2,14 @@
 package config
 
 import (
+	"errors"
 	"fmt"
+	"net/netip"
 	"os"
 	"path/filepath"
 	"runtime"
 	"strconv"
+	"strings"
 	"time"
 )
 
@@ -27,6 +30,14 @@ const (
 	maxPort = 65535
 )
 
+var (
+	errRelativeXDGDataHome = errors.New("XDG_DATA_HOME must be an absolute path")
+	errUnsupportedOS       = errors.New("unsupported operating system")
+	errInvalidPort         = errors.New("PORT must be a whole number")
+	errInvalidProxy        = errors.New(
+		"TRUSTED_PROXIES must list IP addresses and CIDR ranges, comma-separated")
+)
+
 // Config holds configuration for the entire application
 type Config struct {
 	// StateDir is the directory for all application state (database, snapshots)
@@ -35,14 +46,21 @@ type Config struct {
 	// Port is the TCP port the HTTP server listens on
 	Port int
 
+	// TrustedProxies are the reverse proxies whose X-Forwarded-For header
+	// the HTTP server believes
+	TrustedProxies []netip.Prefix
+
 	// MaxRuntime is the maximum runtime (0 = run forever)
 	MaxRuntime time.Duration
 
-	// EnableBatchedDatabaseWrites enables batched database operations for better performance
+	// EnableBatchedDatabaseWrites enables batched database operations for
+	// better performance
 	EnableBatchedDatabaseWrites bool
 
-	// RouteExpirationTimeout is how long a route can go without being refreshed before expiring
+	// RouteExpirationTimeout is how long a route can go without being
-	// Default is 2 hours which is conservative for BGP (typical BGP hold time is 90-180 seconds)
+	// refreshed before expiring
+	// Default is 2 hours which is conservative for BGP (typical BGP hold
+	// time is 90-180 seconds)
 	RouteExpirationTimeout time.Duration
 }
 
@@ -58,12 +76,21 @@ func New() (*Config, error) {
 		return nil, err
 	}
 
+	trustedProxies, err := getTrustedProxies()
+	if err != nil {
+		return nil, err
+	}
+
 	return &Config{
 		StateDir:       stateDir,
 		Port:           port,
-		MaxRuntime:                  0,                                           // Run forever by default
+		TrustedProxies: trustedProxies,
-		EnableBatchedDatabaseWrites: true,                                        // Enable batching by default
+		// Run forever by default
-		RouteExpirationTimeout:      defaultRouteExpirationMinutes * time.Minute, // For active route monitoring
+		MaxRuntime: 0,
+		// Enable batching by default
+		EnableBatchedDatabaseWrites: true,
+		// For active route monitoring
+		RouteExpirationTimeout: defaultRouteExpirationMinutes * time.Minute,
 	}, nil
 }
 
@@ -89,7 +116,7 @@ func getStateDirectory() (string, error) {
 		// the database somewhere unexpected.
 		xdgData := os.Getenv("XDG_DATA_HOME")
 		if xdgData != "" && !filepath.IsAbs(xdgData) {
-			return "", fmt.Errorf("XDG_DATA_HOME must be an absolute path, got %q", xdgData)
+			return "", fmt.Errorf("%w, got %q", errRelativeXDGDataHome, xdgData)
 		}
 
 		// Unix-like: /var/lib/berlin.sneak.app.routewatch if root, else XDG_DATA_HOME
@@ -111,7 +138,7 @@ func getStateDirectory() (string, error) {
 		return filepath.Join(home, ".local", "share", AppIdentifier), nil
 
 	default:
-		return "", fmt.Errorf("unsupported operating system: %s", runtime.GOOS)
+		return "", fmt.Errorf("%w: %s", errUnsupportedOS, runtime.GOOS)
 	}
 }
 
@@ -125,16 +152,50 @@ func getPort() (int, error) {
 	// ParseUint, unlike Atoi, refuses a sign: the health check URL cannot use "+9090"
 	port, err := strconv.ParseUint(value, 10, 0)
 	if err != nil || port < 1 || port > maxPort {
-		return 0, fmt.Errorf("PORT must be a whole number from 1 to %d, got %q", maxPort, value)
+		return 0, fmt.Errorf("%w from 1 to %d, got %q",
+			errInvalidPort, maxPort, value)
 	}
 
 	return int(port), nil
 }
 
+// getTrustedProxies returns the addresses and CIDR ranges listed in
+// TRUSTED_PROXIES, such as "10.0.0.0/8,192.0.2.1", or none when it is not set.
+// An address is returned as a range holding only that address.
+func getTrustedProxies() ([]netip.Prefix, error) {
+	value := os.Getenv("TRUSTED_PROXIES")
+	if value == "" {
+		return nil, nil
+	}
+
+	var proxies []netip.Prefix
+
+	for entry := range strings.SplitSeq(value, ",") {
+		entry = strings.TrimSpace(entry)
+
+		addr, err := netip.ParseAddr(entry)
+		if err == nil {
+			proxies = append(proxies, netip.PrefixFrom(addr, addr.BitLen()))
+
+			continue
+		}
+
+		prefix, err := netip.ParsePrefix(entry)
+		if err != nil {
+			return nil, fmt.Errorf("%w, got %q", errInvalidProxy, entry)
+		}
+
+		proxies = append(proxies, prefix)
+	}
+
+	return proxies, nil
+}
+
 // EnsureDirectories creates all necessary directories if they don't exist
 func (c *Config) EnsureDirectories() error {
 	// Ensure state directory exists
-	if err := os.MkdirAll(c.StateDir, dirPermissions); err != nil {
+	err := os.MkdirAll(c.StateDir, dirPermissions)
+	if err != nil {
 		return fmt.Errorf("failed to create state directory: %w", err)
 	}
 

internal/config/config_internal_test.go

@@ -0,0 +1,118 @@
+package config
+
+import (
+	"net/netip"
+	"runtime"
+	"slices"
+	"testing"
+)
+
+func TestNewReadsPort(t *testing.T) {
+	tests := map[string]int{
+		"":      defaultPort,
+		"1":     1,
+		"9090":  9090,
+		"65535": 65535,
+	}
+
+	for value, want := range tests {
+		t.Run(value, func(t *testing.T) {
+			t.Setenv("PORT", value)
+			t.Setenv("XDG_DATA_HOME", "")
+
+			cfg, err := New()
+			if err != nil {
+				t.Fatalf("New() with PORT=%q: %v", value, err)
+			}
+
+			if cfg.Port != want {
+				t.Errorf("New() with PORT=%q: Port = %d, want %d", value, cfg.Port, want)
+			}
+		})
+	}
+}
+
+func TestNewRefusesInvalidPort(t *testing.T) {
+	for _, value := range []string{"0", "65536", "-1", "+9090", "http", "80.5"} {
+		t.Run(value, func(t *testing.T) {
+			t.Setenv("PORT", value)
+			t.Setenv("XDG_DATA_HOME", "")
+
+			_, err := New()
+			if err == nil {
+				t.Errorf("New() with PORT=%q returned no error", value)
+			}
+		})
+	}
+}
+
+func TestNewReadsTrustedProxies(t *testing.T) {
+	t.Setenv("PORT", "")
+	t.Setenv("XDG_DATA_HOME", "")
+
+	t.Setenv("TRUSTED_PROXIES", "")
+
+	cfg, err := New()
+	if err != nil {
+		t.Fatalf("New() without TRUSTED_PROXIES: %v", err)
+	}
+
+	if len(cfg.TrustedProxies) != 0 {
+		t.Errorf("New() without TRUSTED_PROXIES: TrustedProxies = %v, want none",
+			cfg.TrustedProxies)
+	}
+
+	t.Setenv("TRUSTED_PROXIES", "10.0.0.0/8, 192.0.2.1,2001:db8::/32")
+
+	cfg, err = New()
+	if err != nil {
+		t.Fatalf("New() with TRUSTED_PROXIES: %v", err)
+	}
+
+	want := []netip.Prefix{
+		netip.MustParsePrefix("10.0.0.0/8"),
+		netip.MustParsePrefix("192.0.2.1/32"),
+		netip.MustParsePrefix("2001:db8::/32"),
+	}
+	if !slices.Equal(cfg.TrustedProxies, want) {
+		t.Errorf("TrustedProxies = %v, want %v", cfg.TrustedProxies, want)
+	}
+}
+
+func TestNewRefusesInvalidTrustedProxies(t *testing.T) {
+	invalid := []string{"proxy.example", "10.0.0.0/33", "192.0.2.1:80", "192.0.2.1,"}
+	for _, value := range invalid {
+		t.Run(value, func(t *testing.T) {
+			t.Setenv("PORT", "")
+			t.Setenv("XDG_DATA_HOME", "")
+			t.Setenv("TRUSTED_PROXIES", value)
+
+			_, err := New()
+			if err == nil {
+				t.Errorf("New() with TRUSTED_PROXIES=%q returned no error", value)
+			}
+		})
+	}
+}
+
+func TestNewRefusesRelativeXDGDataHome(t *testing.T) {
+	if runtime.GOOS == "darwin" {
+		t.Skip("macOS does not read XDG_DATA_HOME")
+	}
+
+	t.Setenv("PORT", "")
+
+	t.Setenv("XDG_DATA_HOME", "relative/path")
+
+	_, err := New()
+	if err == nil {
+		t.Error("New() with a relative XDG_DATA_HOME returned no error")
+	}
+
+	t.Setenv("XDG_DATA_HOME", "/var/lib")
+
+	_, err = New()
+	if err != nil {
+		t.Errorf("New() with XDG_DATA_HOME=/var/lib: %v", err)
+	}
+}

internal/config/config_test.go

@@ -1,62 +0,0 @@
-package config
-
-import (
-	"runtime"
-	"testing"
-)
-
-func TestNewReadsPort(t *testing.T) {
-	tests := map[string]int{
-		"":      defaultPort,
-		"1":     1,
-		"9090":  9090,
-		"65535": 65535,
-	}
-
-	for value, want := range tests {
-		t.Run(value, func(t *testing.T) {
-			t.Setenv("PORT", value)
-			t.Setenv("XDG_DATA_HOME", "")
-
-			cfg, err := New()
-			if err != nil {
-				t.Fatalf("New() with PORT=%q: %v", value, err)
-			}
-
-			if cfg.Port != want {
-				t.Errorf("New() with PORT=%q: Port = %d, want %d", value, cfg.Port, want)
-			}
-		})
-	}
-}
-
-func TestNewRefusesInvalidPort(t *testing.T) {
-	for _, value := range []string{"0", "65536", "-1", "+9090", "http", "80.5"} {
-		t.Run(value, func(t *testing.T) {
-			t.Setenv("PORT", value)
-			t.Setenv("XDG_DATA_HOME", "")
-
-			if _, err := New(); err == nil {
-				t.Errorf("New() with PORT=%q returned no error", value)
-			}
-		})
-	}
-}
-
-func TestNewRefusesRelativeXDGDataHome(t *testing.T) {
-	if runtime.GOOS == "darwin" {
-		t.Skip("macOS does not read XDG_DATA_HOME")
-	}
-
-	t.Setenv("PORT", "")
-
-	t.Setenv("XDG_DATA_HOME", "relative/path")
-	if _, err := New(); err == nil {
-		t.Error("New() with a relative XDG_DATA_HOME returned no error")
-	}
-
-	t.Setenv("XDG_DATA_HOME", "/var/lib")
-	if _, err := New(); err != nil {
-		t.Errorf("New() with XDG_DATA_HOME=/var/lib: %v", err)
-	}
-}

internal/database/counts.go

@@ -33,8 +33,10 @@ type liveCounts struct {
 
 // seed sets every count to the value read from the tables at startup. It runs
 // before any writer, so it needs no coordination with the adjust methods.
-func (c *liveCounts) seed(asns, prefixesV4, prefixesV6, peerings, peers, routesV4, routesV6 int,
+func (c *liveCounts) seed(
-	distributionV4, distributionV6 []PrefixDistribution) {
+	asns, prefixesV4, prefixesV6, peerings, peers, routesV4, routesV6 int,
+	distributionV4, distributionV6 []PrefixDistribution,
+) {
 	c.mu.Lock()
 	defer c.mu.Unlock()
 
@@ -45,9 +47,11 @@ func (c *liveCounts) seed(asns, prefixesV4, prefixesV6, peerings, peers, routesV
 	c.peers = peers
 	c.routesV4 = routesV4
 	c.routesV6 = routesV6
+
 	for _, entry := range distributionV4 {
 		addAtMaskLength(c.distributionV4[:], entry.MaskLength, entry.Count)
 	}
+
 	for _, entry := range distributionV6 {
 		addAtMaskLength(c.distributionV6[:], entry.MaskLength, entry.Count)
 	}
@@ -102,6 +106,7 @@ func (c *liveCounts) addToDistribution(maskLengthsV4, maskLengthsV6 []int, n int
 	for _, maskLength := range maskLengthsV4 {
 		addAtMaskLength(c.distributionV4[:], maskLength, n)
 	}
+
 	for _, maskLength := range maskLengthsV6 {
 		addAtMaskLength(c.distributionV6[:], maskLength, n)
 	}
@@ -138,9 +143,11 @@ func (c *liveCounts) fill(s *Stats) {
 // ascending order, the way the distribution query returns them.
 func distributionList(counts []int) []PrefixDistribution {
 	var list []PrefixDistribution
+
 	for maskLength, count := range counts {
 		if count > 0 {
-			list = append(list, PrefixDistribution{MaskLength: maskLength, Count: count})
+			list = append(list,
+				PrefixDistribution{MaskLength: maskLength, Count: count})
 		}
 	}
 
@@ -152,7 +159,8 @@ func distributionList(counts []int) []PrefixDistribution {
 func (d *Database) countRows(ctx context.Context, table string) (int, error) {
 	var n int
 	// table is one of a fixed set of literals below, never external input.
-	if err := d.db.QueryRowContext(ctx, "SELECT COUNT(*) FROM "+table).Scan(&n); err != nil {
+	err := d.db.QueryRowContext(ctx, "SELECT COUNT(*) FROM "+table).Scan(&n)
+	if err != nil {
 		return 0, fmt.Errorf("failed to count %s: %w", table, err)
 	}
 
@@ -167,30 +175,37 @@ func (d *Database) seedCounts(ctx context.Context) error {
 	if err != nil {
 		return err
 	}
+
 	prefixesV4, err := d.countRows(ctx, "prefixes_v4")
 	if err != nil {
 		return err
 	}
+
 	prefixesV6, err := d.countRows(ctx, "prefixes_v6")
 	if err != nil {
 		return err
 	}
+
 	peerings, err := d.countRows(ctx, "peerings")
 	if err != nil {
 		return err
 	}
+
 	peers, err := d.countRows(ctx, "bgp_peers")
 	if err != nil {
 		return err
 	}
-	routesV4, err := d.countRows(ctx, "live_routes_v4")
+
+	routesV4, err := d.countRows(ctx, tableLiveRoutesV4)
 	if err != nil {
 		return err
 	}
-	routesV6, err := d.countRows(ctx, "live_routes_v6")
+
+	routesV6, err := d.countRows(ctx, tableLiveRoutesV6)
 	if err != nil {
 		return err
 	}
+
 	distributionV4, distributionV6, err := d.GetPrefixDistributionContext(ctx)
 	if err != nil {
 		return err

internal/database/counts_internal_test.go

@@ -2,6 +2,7 @@ package database
 
 import (
 	"context"
+	"fmt"
 	"slices"
 	"sync"
 	"testing"
@@ -12,6 +13,12 @@ import (
 	"github.com/google/uuid"
 )
 
+// testV4PeerIP is the peer the IPv4 routes in these tests are learned from.
+const testV4PeerIP = "192.0.2.1"
+
+// testV6Prefix is the prefix of the IPv6 routes in these tests.
+const testV6Prefix = "2001:db8::/32"
+
 // mkV4Route builds an IPv4 live route with its mask length taken from the
 // prefix.
 func mkV4Route(t *testing.T, prefix string, asn int, ts time.Time) *LiveRoute {
@@ -23,12 +30,11 @@ func mkV4Route(t *testing.T, prefix string, asn int, ts time.Time) *LiveRoute {
 	}
 
 	return &LiveRoute{
-		ID:          uuid.New(),
 		Prefix:      prefix,
 		MaskLength:  maskLength,
 		IPVersion:   ipVersionV4,
 		OriginASN:   asn,
-		PeerIP:      "192.0.2.1",
+		PeerIP:      testV4PeerIP,
 		ASPath:      []int{asn},
 		NextHop:     "192.0.2.254",
 		LastUpdated: ts,
@@ -38,7 +44,6 @@ func mkV4Route(t *testing.T, prefix string, asn int, ts time.Time) *LiveRoute {
 // mkV6Route builds an IPv6 live route.
 func mkV6Route(prefix string, asn int, ts time.Time) *LiveRoute {
 	return &LiveRoute{
-		ID:          uuid.New(),
 		Prefix:      prefix,
 		MaskLength:  32,
 		IPVersion:   ipVersionV6,
@@ -56,6 +61,8 @@ func mkV6Route(prefix string, asn int, ts time.Time) *LiveRoute {
 // deleted. These counts are what /api/v1/stats reports; before this change the
 // endpoint recomputed them with a COUNT(*) over each table on every request.
 func TestLiveCountsTrackWritesInRealtime(t *testing.T) {
+	t.Parallel()
+
 	cfg := &config.Config{StateDir: t.TempDir()}
 
 	db, err := New(cfg, logger.New())
@@ -70,6 +77,7 @@ func TestLiveCountsTrackWritesInRealtime(t *testing.T) {
 	if err != nil {
 		t.Fatalf("GetStatsContext on empty database: %v", err)
 	}
+
 	if empty.ASNs != 0 || empty.Prefixes != 0 || empty.Peerings != 0 ||
 		empty.Peers != 0 || empty.LiveRoutes != 0 {
 		t.Fatalf("empty database counts nonzero: %+v", empty)
@@ -77,30 +85,16 @@ func TestLiveCountsTrackWritesInRealtime(t *testing.T) {
 
 	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
 
-	if err := db.GetOrCreateASNBatch(map[int]time.Time{64500: ts, 64501: ts}); err != nil {
+	writeASNsPrefixesAndPeers(t, db, ts)
-		t.Fatalf("GetOrCreateASNBatch: %v", err)
-	}
-	if err := db.UpdatePrefixesBatch(map[string]time.Time{
-		"198.51.100.0/24": ts,
-		"2001:db8::/32":   ts,
-	}); err != nil {
-		t.Fatalf("UpdatePrefixesBatch: %v", err)
-	}
-	if err := db.UpdatePeerBatch(map[string]PeerUpdate{
-		"192.0.2.1": {PeerIP: "192.0.2.1", PeerASN: 64500, MessageType: "UPDATE", Timestamp: ts},
-	}); err != nil {
-		t.Fatalf("UpdatePeerBatch: %v", err)
-	}
-	if err := db.RecordPeering(64500, 64501, ts); err != nil {
-		t.Fatalf("RecordPeering: %v", err)
-	}
 
 	routes := []*LiveRoute{
 		mkV4Route(t, "198.51.100.0/24", 64500, ts),
 		mkV4Route(t, "203.0.113.0/24", 64501, ts.Add(time.Minute)),
-		mkV6Route("2001:db8::/32", 64502, ts.Add(2*time.Minute)),
+		mkV6Route(testV6Prefix, 64502, ts.Add(2*time.Minute)),
 	}
-	if err := db.UpsertLiveRouteBatch(routes); err != nil {
+
+	err = db.UpsertLiveRouteBatch(routes)
+	if err != nil {
 		t.Fatalf("UpsertLiveRouteBatch: %v", err)
 	}
 
@@ -108,44 +102,89 @@ func TestLiveCountsTrackWritesInRealtime(t *testing.T) {
 	if err != nil {
 		t.Fatalf("GetStatsContext: %v", err)
 	}
+
 	assertCounts(t, "after inserts", stats, wantCounts{
 		asns: 2, prefixes: 2, peerings: 1, peers: 1,
 		ipv4Routes: 2, ipv6Routes: 1, liveRoutes: 3,
 	})
 
 	// Re-announcing the same routes is an update, not an insert: counts hold.
-	if err := db.UpsertLiveRouteBatch(routes); err != nil {
+	err = db.UpsertLiveRouteBatch(routes)
+	if err != nil {
 		t.Fatalf("UpsertLiveRouteBatch (re-announce): %v", err)
 	}
+
 	stats, err = db.GetStatsContext(ctx)
 	if err != nil {
 		t.Fatalf("GetStatsContext: %v", err)
 	}
+
 	assertCounts(t, "after re-announce", stats, wantCounts{
 		asns: 2, prefixes: 2, peerings: 1, peers: 1,
 		ipv4Routes: 2, ipv6Routes: 1, liveRoutes: 3,
 	})
 
 	// A withdrawal removes one route.
-	if err := db.DeleteLiveRouteBatch([]LiveRouteDeletion{
+	err = db.DeleteLiveRouteBatch([]LiveRouteDeletion{
-		{Prefix: "203.0.113.0/24", OriginASN: 64501, PeerIP: "192.0.2.1", IPVersion: ipVersionV4},
+		{Prefix: "203.0.113.0/24", PeerIP: testV4PeerIP, IPVersion: ipVersionV4},
-	}); err != nil {
+	})
+	if err != nil {
 		t.Fatalf("DeleteLiveRouteBatch: %v", err)
 	}
+
 	stats, err = db.GetStatsContext(ctx)
 	if err != nil {
 		t.Fatalf("GetStatsContext: %v", err)
 	}
+
 	assertCounts(t, "after delete", stats, wantCounts{
 		asns: 2, prefixes: 2, peerings: 1, peers: 1,
 		ipv4Routes: 1, ipv6Routes: 1, liveRoutes: 2,
 	})
 }
 
+// writeASNsPrefixesAndPeers writes two ASNs, two prefixes, one peer and one
+// peering: the rows besides live routes that the stats count.
+func writeASNsPrefixesAndPeers(t *testing.T, db *Database, ts time.Time) {
+	t.Helper()
+
+	seen := SeenTimes{First: ts, Last: ts}
+
+	err := db.GetOrCreateASNBatch(map[int]SeenTimes{64500: seen, 64501: seen})
+	if err != nil {
+		t.Fatalf("GetOrCreateASNBatch: %v", err)
+	}
+
+	err = db.UpdatePrefixesBatch(map[string]time.Time{
+		"198.51.100.0/24": ts,
+		testV6Prefix:      ts,
+	})
+	if err != nil {
+		t.Fatalf("UpdatePrefixesBatch: %v", err)
+	}
+
+	err = db.UpdatePeerBatch(map[string]PeerUpdate{
+		testV4PeerIP: {
+			PeerIP: testV4PeerIP, PeerASN: 64500,
+			MessageType: "UPDATE", Seen: seen,
+		},
+	})
+	if err != nil {
+		t.Fatalf("UpdatePeerBatch: %v", err)
+	}
+
+	err = db.RecordPeering(64500, 64501, seen)
+	if err != nil {
+		t.Fatalf("RecordPeering: %v", err)
+	}
+}
+
 // TestLiveCountsSeededFromDatabaseAtStartup writes rows, reopens the same
 // database file, and checks the counts and the prefix distribution come back
 // from the seed scan rather than starting at zero.
 func TestLiveCountsSeededFromDatabaseAtStartup(t *testing.T) {
+	t.Parallel()
+
 	cfg := &config.Config{StateDir: t.TempDir()}
 
 	db, err := New(cfg, logger.New())
@@ -154,16 +193,25 @@ func TestLiveCountsSeededFromDatabaseAtStartup(t *testing.T) {
 	}
 
 	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
-	if err := db.GetOrCreateASNBatch(map[int]time.Time{64500: ts, 64501: ts, 64502: ts}); err != nil {
+
+	seen := SeenTimes{First: ts, Last: ts}
+
+	err = db.GetOrCreateASNBatch(
+		map[int]SeenTimes{64500: seen, 64501: seen, 64502: seen})
+	if err != nil {
 		t.Fatalf("GetOrCreateASNBatch: %v", err)
 	}
-	if err := db.UpsertLiveRouteBatch([]*LiveRoute{
+
+	err = db.UpsertLiveRouteBatch([]*LiveRoute{
 		mkV4Route(t, "198.51.100.0/24", 64500, ts),
-		mkV6Route("2001:db8::/32", 64502, ts),
+		mkV6Route(testV6Prefix, 64502, ts),
-	}); err != nil {
+	})
+	if err != nil {
 		t.Fatalf("UpsertLiveRouteBatch: %v", err)
 	}
-	if err := db.Close(); err != nil {
+
+	err = db.Close()
+	if err != nil {
 		t.Fatalf("Close: %v", err)
 	}
 
@@ -177,13 +225,16 @@ func TestLiveCountsSeededFromDatabaseAtStartup(t *testing.T) {
 	if err != nil {
 		t.Fatalf("GetStatsContext after reopen: %v", err)
 	}
+
 	if stats.ASNs != 3 {
 		t.Errorf("seeded ASNs = %d, want 3", stats.ASNs)
 	}
+
 	if stats.IPv4Routes != 1 || stats.IPv6Routes != 1 || stats.LiveRoutes != 2 {
 		t.Errorf("seeded routes = (v4 %d, v6 %d, total %d), want (1, 1, 2)",
 			stats.IPv4Routes, stats.IPv6Routes, stats.LiveRoutes)
 	}
+
 	assertDistribution(t, "seeded IPv4 distribution", stats.IPv4PrefixDistribution,
 		[]PrefixDistribution{{MaskLength: 24, Count: 1}})
 	assertDistribution(t, "seeded IPv6 distribution", stats.IPv6PrefixDistribution,
@@ -196,26 +247,91 @@ func TestLiveCountsSeededFromDatabaseAtStartup(t *testing.T) {
 // tables. The steps run once through the batch methods the prefix handler uses
 // and once through the single-route methods.
 func TestPrefixDistributionTracksWrites(t *testing.T) {
+	t.Parallel()
+
+	steps := prefixDistributionSteps(t)
+
+	for _, batch := range []bool{true, false} {
+		name := "single-route writes"
+		if batch {
+			name = "batch writes"
+		}
+
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			db, err := New(&config.Config{StateDir: t.TempDir()}, logger.New())
+			if err != nil {
+				t.Fatalf("failed to create database: %v", err)
+			}
+			defer func() { _ = db.Close() }()
+
+			ctx := context.Background()
+
+			for _, step := range steps {
+				err := announceRoutes(db, batch, step.announce)
+				if err != nil {
+					t.Fatalf("%s: announce: %v", step.name, err)
+				}
+
+				err = withdrawRoutes(db, batch, step.withdraw)
+				if err != nil {
+					t.Fatalf("%s: withdraw: %v", step.name, err)
+				}
+
+				stats, err := db.GetStatsContext(ctx)
+				if err != nil {
+					t.Fatalf("%s: GetStatsContext: %v", step.name, err)
+				}
+
+				queryV4, queryV6, err := db.GetPrefixDistributionContext(ctx)
+				if err != nil {
+					t.Fatalf("%s: GetPrefixDistributionContext: %v", step.name, err)
+				}
+
+				assertDistribution(t, step.name+": IPv4 distribution",
+					stats.IPv4PrefixDistribution, step.wantV4)
+				assertDistribution(t, step.name+": IPv6 distribution",
+					stats.IPv6PrefixDistribution, step.wantV6)
+				assertDistribution(t, step.name+": IPv4 distribution query",
+					queryV4, step.wantV4)
+				assertDistribution(t, step.name+": IPv6 distribution query",
+					queryV6, step.wantV6)
+			}
+		})
+	}
+}
+
+// prefixDistributionStep is one step of TestPrefixDistributionTracksWrites:
+// the routes it announces, then the routes it withdraws, then the
+// distributions both the stats and the distribution query must report.
+type prefixDistributionStep struct {
+	name     string
+	announce []*LiveRoute
+	withdraw []*LiveRoute
+	wantV4   []PrefixDistribution
+	wantV6   []PrefixDistribution
+}
+
+// prefixDistributionSteps returns the steps of
+// TestPrefixDistributionTracksWrites, in the order they run.
+func prefixDistributionSteps(t *testing.T) []prefixDistributionStep {
+	t.Helper()
+
 	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
 	shared := mkV4Route(t, "198.51.100.0/24", 64500, ts)
 	sharedSecondPeer := mkV4Route(t, "198.51.100.0/24", 64500, ts)
 	sharedSecondPeer.PeerIP = "192.0.2.2"
 	other := mkV4Route(t, "203.0.113.0/24", 64501, ts)
 	wide := mkV4Route(t, "172.16.0.0/16", 64502, ts)
-	v6 := mkV6Route("2001:db8::/32", 64503, ts)
+	v6 := mkV6Route(testV6Prefix, 64503, ts)
-	v6SecondPeer := mkV6Route("2001:db8::/32", 64503, ts)
+	v6SecondPeer := mkV6Route(testV6Prefix, 64503, ts)
 	v6SecondPeer.PeerIP = "2001:db8::2"
 
 	all := []PrefixDistribution{{MaskLength: 16, Count: 1}, {MaskLength: 24, Count: 2}}
 	v6Only := []PrefixDistribution{{MaskLength: 32, Count: 1}}
 
-	steps := []struct {
+	return []prefixDistributionStep{
-		name     string
-		announce []*LiveRoute
-		withdraw []*LiveRoute
-		wantV4   []PrefixDistribution
-		wantV6   []PrefixDistribution
-	}{
 		{
 			name:     "new routes",
 			announce: []*LiveRoute{shared, other, wide, v6},
@@ -239,7 +355,9 @@ func TestPrefixDistributionTracksWrites(t *testing.T) {
 		{
 			name:     "withdrawal of the last route for a prefix",
 			withdraw: []*LiveRoute{sharedSecondPeer, v6SecondPeer},
-			wantV4:   []PrefixDistribution{{MaskLength: 16, Count: 1}, {MaskLength: 24, Count: 1}},
+			wantV4: []PrefixDistribution{
+				{MaskLength: 16, Count: 1}, {MaskLength: 24, Count: 1},
+			},
 		},
 		{
 			name:     "withdrawal of every remaining route",
@@ -270,45 +388,6 @@ func TestPrefixDistributionTracksWrites(t *testing.T) {
 			wantV6:   v6Only,
 		},
 	}
-
-	for _, batch := range []bool{true, false} {
-		name := "single-route writes"
-		if batch {
-			name = "batch writes"
-		}
-
-		t.Run(name, func(t *testing.T) {
-			db, err := New(&config.Config{StateDir: t.TempDir()}, logger.New())
-			if err != nil {
-				t.Fatalf("failed to create database: %v", err)
-			}
-			defer func() { _ = db.Close() }()
-
-			ctx := context.Background()
-			for _, step := range steps {
-				if err := announceRoutes(db, batch, step.announce); err != nil {
-					t.Fatalf("%s: announce: %v", step.name, err)
-				}
-				if err := withdrawRoutes(db, batch, step.withdraw); err != nil {
-					t.Fatalf("%s: withdraw: %v", step.name, err)
-				}
-
-				stats, err := db.GetStatsContext(ctx)
-				if err != nil {
-					t.Fatalf("%s: GetStatsContext: %v", step.name, err)
-				}
-				queryV4, queryV6, err := db.GetPrefixDistributionContext(ctx)
-				if err != nil {
-					t.Fatalf("%s: GetPrefixDistributionContext: %v", step.name, err)
-				}
-
-				assertDistribution(t, step.name+": IPv4 distribution", stats.IPv4PrefixDistribution, step.wantV4)
-				assertDistribution(t, step.name+": IPv6 distribution", stats.IPv6PrefixDistribution, step.wantV6)
-				assertDistribution(t, step.name+": IPv4 distribution query", queryV4, step.wantV4)
-				assertDistribution(t, step.name+": IPv6 distribution query", queryV6, step.wantV6)
-			}
-		})
-	}
 }
 
 // announceRoutes writes routes in one UpsertLiveRouteBatch, or with one
@@ -317,8 +396,10 @@ func announceRoutes(db *Database, batch bool, routes []*LiveRoute) error {
 	if batch {
 		return db.UpsertLiveRouteBatch(routes)
 	}
+
 	for _, route := range routes {
-		if err := db.UpsertLiveRoute(route); err != nil {
+		err := db.UpsertLiveRoute(route)
+		if err != nil {
 			return err
 		}
 	}
@@ -327,13 +408,12 @@ func announceRoutes(db *Database, batch bool, routes []*LiveRoute) error {
 }
 
 // withdrawRoutes removes routes in one DeleteLiveRouteBatch, or with one
-// DeleteLiveRoute each. It names each route by prefix and peer only, with no
+// DeleteLiveRoute each.
-// origin ASN, as a withdrawal from the feed does when its message carries no AS
-// path.
 func withdrawRoutes(db *Database, batch bool, routes []*LiveRoute) error {
 	if !batch {
 		for _, route := range routes {
-			if err := db.DeleteLiveRoute(route.Prefix, 0, route.PeerIP); err != nil {
+			err := db.DeleteLiveRoute(route.Prefix, route.PeerIP)
+			if err != nil {
 				return err
 			}
 		}
@@ -351,11 +431,239 @@ func withdrawRoutes(db *Database, batch bool, routes []*LiveRoute) error {
 	return db.DeleteLiveRouteBatch(deletions)
 }
 
+// TestDeleteLiveRoutesOfPeer gives two peers IPv4 and IPv6 live routes, some
+// for the same prefixes, and deletes all of one peer's routes. The other peer's
+// routes stay, and the route counts and prefix distribution held in memory
+// equal what is read from the route tables.
+func TestDeleteLiveRoutesOfPeer(t *testing.T) {
+	t.Parallel()
+
+	const (
+		peerIP      = "192.0.2.1"
+		otherPeerIP = "192.0.2.2"
+		sharedV4    = "198.51.100.0/24"
+		sharedV6    = "2001:db8::/32"
+	)
+
+	db, err := New(&config.Config{StateDir: t.TempDir()}, logger.New())
+	if err != nil {
+		t.Fatalf("failed to create database: %v", err)
+	}
+	defer func() { _ = db.Close() }()
+
+	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
+	ofPeer := func(peer string, route *LiveRoute) *LiveRoute {
+		route.PeerIP = peer
+
+		return route
+	}
+	onlyV6 := mkV6Route("2001:db8:1::/48", 64503, ts)
+	onlyV6.MaskLength = 48
+
+	err = db.UpsertLiveRouteBatch([]*LiveRoute{
+		ofPeer(peerIP, mkV4Route(t, sharedV4, 64500, ts)),
+		ofPeer(otherPeerIP, mkV4Route(t, sharedV4, 64500, ts)),
+		ofPeer(peerIP, mkV4Route(t, "203.0.113.0/24", 64501, ts)),
+		ofPeer(peerIP, mkV6Route(sharedV6, 64502, ts)),
+		ofPeer(otherPeerIP, mkV6Route(sharedV6, 64502, ts)),
+		ofPeer(peerIP, onlyV6),
+	})
+	if err != nil {
+		t.Fatalf("UpsertLiveRouteBatch: %v", err)
+	}
+
+	// peerIP has two IPv4 and two IPv6 routes; a second delete finds none.
+	for _, want := range []int{4, 0} {
+		deleted, err := db.DeleteLiveRoutesOfPeer(peerIP)
+		if err != nil {
+			t.Fatalf("DeleteLiveRoutesOfPeer: %v", err)
+		}
+
+		if deleted != want {
+			t.Errorf("DeleteLiveRoutesOfPeer deleted %d routes, want %d",
+				deleted, want)
+		}
+	}
+
+	for _, prefix := range []string{sharedV4, sharedV6} {
+		routes, _, err := db.GetPrefixDetailsContext(context.Background(), prefix)
+		if err != nil {
+			t.Fatalf("GetPrefixDetailsContext(%s): %v", prefix, err)
+		}
+
+		if len(routes) != 1 || routes[0].PeerIP != otherPeerIP {
+			t.Errorf("live routes for %s = %+v, want only the one from %s",
+				prefix, routes, otherPeerIP)
+		}
+	}
+
+	checkOneLiveRoutePerIPVersion(t, db)
+}
+
+// TestDeleteLiveRoutesOfPeerDeletesInChunks gives a peer one IPv4 route more
+// than a chunk holds, two of them for prefixes another peer also has. One chunk
+// deletes peerRouteChunkSize of the peer's routes and commits, leaving the last
+// one. The routes are written again, and DeleteLiveRoutesOfPeer deletes all of
+// them, a chunk and then the last one. After each step the route count and
+// prefix distribution held in memory and those read from the route table are
+// the ones wanted.
+func TestDeleteLiveRoutesOfPeerDeletesInChunks(t *testing.T) {
+	t.Parallel()
+
+	const (
+		peerIP      = "192.0.2.1"
+		otherPeerIP = "192.0.2.2"
+		// How many of the peer's prefixes the other peer also has, the first
+		// ones in prefix order.
+		sharedPrefixes = 2
+	)
+
+	db, err := New(&config.Config{StateDir: t.TempDir()}, logger.New())
+	if err != nil {
+		t.Fatalf("failed to create database: %v", err)
+	}
+	defer func() { _ = db.Close() }()
+
+	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
+
+	peerRoutes := make([]*LiveRoute, 0, peerRouteChunkSize+1)
+	otherPeerRoutes := make([]*LiveRoute, 0, sharedPrefixes)
+
+	for i := range peerRouteChunkSize + 1 {
+		prefix := fmt.Sprintf("10.%d.%d.0/24", i/256, i%256)
+
+		route := mkV4Route(t, prefix, 64500, ts)
+		route.PeerIP = peerIP
+		peerRoutes = append(peerRoutes, route)
+
+		if i < sharedPrefixes {
+			route := mkV4Route(t, prefix, 64500, ts)
+			route.PeerIP = otherPeerIP
+			otherPeerRoutes = append(otherPeerRoutes, route)
+		}
+	}
+
+	err = db.UpsertLiveRouteBatch(slices.Concat(peerRoutes, otherPeerRoutes))
+	if err != nil {
+		t.Fatalf("UpsertLiveRouteBatch: %v", err)
+	}
+
+	deletedV4, deletedV6, err := db.deletePeerRouteChunk(peerIP)
+	if err != nil {
+		t.Fatalf("deletePeerRouteChunk: %v", err)
+	}
+
+	if deletedV4 != peerRouteChunkSize || deletedV6 != 0 {
+		t.Errorf("deletePeerRouteChunk deleted (%d, %d) routes, want (%d, 0)",
+			deletedV4, deletedV6, peerRouteChunkSize)
+	}
+
+	// The shared prefixes went with the chunk, so each route left has a
+	// prefix of its own.
+	checkV4RoutesOfOwnPrefixes(t, db, "after one chunk", 1+sharedPrefixes)
+
+	err = db.UpsertLiveRouteBatch(peerRoutes)
+	if err != nil {
+		t.Fatalf("UpsertLiveRouteBatch: %v", err)
+	}
+
+	deleted, err := db.DeleteLiveRoutesOfPeer(peerIP)
+	if err != nil {
+		t.Fatalf("DeleteLiveRoutesOfPeer: %v", err)
+	}
+
+	if deleted != len(peerRoutes) {
+		t.Errorf("DeleteLiveRoutesOfPeer deleted %d routes, want %d",
+			deleted, len(peerRoutes))
+	}
+
+	checkV4RoutesOfOwnPrefixes(t, db, "after DeleteLiveRoutesOfPeer",
+		sharedPrefixes)
+}
+
+// checkV4RoutesOfOwnPrefixes checks that the database has wantRoutes IPv4 live
+// routes, each for a /24 no other live route has, both as counted in memory and
+// as read from the route table. when says which step of the test checks.
+func checkV4RoutesOfOwnPrefixes(
+	t *testing.T, db *Database, when string, wantRoutes int,
+) {
+	t.Helper()
+
+	ctx := context.Background()
+
+	stats, err := db.GetStatsContext(ctx)
+	if err != nil {
+		t.Fatalf("GetStatsContext: %v", err)
+	}
+
+	tableRoutes, _, err := db.GetLiveRouteCountsContext(ctx)
+	if err != nil {
+		t.Fatalf("GetLiveRouteCountsContext: %v", err)
+	}
+
+	if stats.IPv4Routes != wantRoutes || tableRoutes != wantRoutes {
+		t.Errorf("%s: IPv4 live routes: %d in memory, %d in the table, want %d",
+			when, stats.IPv4Routes, tableRoutes, wantRoutes)
+	}
+
+	queryDistribution, _, err := db.GetPrefixDistributionContext(ctx)
+	if err != nil {
+		t.Fatalf("GetPrefixDistributionContext: %v", err)
+	}
+
+	want := []PrefixDistribution{{MaskLength: 24, Count: wantRoutes}}
+	assertDistribution(t, when+": IPv4 distribution",
+		stats.IPv4PrefixDistribution, want)
+	assertDistribution(t, when+": IPv4 distribution query",
+		queryDistribution, want)
+}
+
+// checkOneLiveRoutePerIPVersion checks that the database has one IPv4 live
+// route, for a /24, and one IPv6 live route, for a /32, both as counted in
+// memory and as read from the route tables.
+func checkOneLiveRoutePerIPVersion(t *testing.T, db *Database) {
+	t.Helper()
+
+	ctx := context.Background()
+
+	stats, err := db.GetStatsContext(ctx)
+	if err != nil {
+		t.Fatalf("GetStatsContext: %v", err)
+	}
+
+	tableV4, tableV6, err := db.GetLiveRouteCountsContext(ctx)
+	if err != nil {
+		t.Fatalf("GetLiveRouteCountsContext: %v", err)
+	}
+
+	if stats.IPv4Routes != 1 || tableV4 != 1 || stats.IPv6Routes != 1 ||
+		tableV6 != 1 {
+		t.Errorf("live routes (IPv4, IPv6): (%d, %d) in memory, (%d, %d) in "+
+			"the tables, want (1, 1)",
+			stats.IPv4Routes, stats.IPv6Routes, tableV4, tableV6)
+	}
+
+	queryV4, queryV6, err := db.GetPrefixDistributionContext(ctx)
+	if err != nil {
+		t.Fatalf("GetPrefixDistributionContext: %v", err)
+	}
+
+	wantV4 := []PrefixDistribution{{MaskLength: 24, Count: 1}}
+	wantV6 := []PrefixDistribution{{MaskLength: 32, Count: 1}}
+
+	assertDistribution(t, "IPv4 distribution", stats.IPv4PrefixDistribution, wantV4)
+	assertDistribution(t, "IPv6 distribution", stats.IPv6PrefixDistribution, wantV6)
+	assertDistribution(t, "IPv4 distribution query", queryV4, wantV4)
+	assertDistribution(t, "IPv6 distribution query", queryV6, wantV6)
+}
+
 // TestStatsRouteTimestamps checks the oldest/newest route timestamps are read
 // from the right rows across both tables and parse into time.Time. The old
 // MIN/MAX union query read its result into *time.Time, which the driver could
 // not parse, so it logged a warning every call and left both timestamps nil.
 func TestStatsRouteTimestamps(t *testing.T) {
+	t.Parallel()
+
 	cfg := &config.Config{StateDir: t.TempDir()}
 
 	db, err := New(cfg, logger.New())
@@ -370,6 +678,7 @@ func TestStatsRouteTimestamps(t *testing.T) {
 	if err != nil {
 		t.Fatalf("GetStatsContext on empty database: %v", err)
 	}
+
 	if empty.OldestRoute != nil || empty.NewestRoute != nil {
 		t.Fatalf("empty database timestamps = (%v, %v), want (nil, nil)",
 			empty.OldestRoute, empty.NewestRoute)
@@ -378,11 +687,13 @@ func TestStatsRouteTimestamps(t *testing.T) {
 	base := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
 	oldest := base
 	newest := base.Add(2 * time.Minute)
-	if err := db.UpsertLiveRouteBatch([]*LiveRoute{
+
+	err = db.UpsertLiveRouteBatch([]*LiveRoute{
 		mkV4Route(t, "198.51.100.0/24", 64500, base.Add(time.Minute)),
 		mkV4Route(t, "203.0.113.0/24", 64501, oldest),
-		mkV6Route("2001:db8::/32", 64502, newest),
+		mkV6Route(testV6Prefix, 64502, newest),
-	}); err != nil {
+	})
+	if err != nil {
 		t.Fatalf("UpsertLiveRouteBatch: %v", err)
 	}
 
@@ -390,9 +701,11 @@ func TestStatsRouteTimestamps(t *testing.T) {
 	if err != nil {
 		t.Fatalf("GetStatsContext: %v", err)
 	}
+
 	if stats.OldestRoute == nil || !stats.OldestRoute.Equal(oldest) {
 		t.Errorf("OldestRoute = %v, want %v", stats.OldestRoute, oldest)
 	}
+
 	if stats.NewestRoute == nil || !stats.NewestRoute.Equal(newest) {
 		t.Errorf("NewestRoute = %v, want %v", stats.NewestRoute, newest)
 	}
@@ -401,6 +714,8 @@ func TestStatsRouteTimestamps(t *testing.T) {
 // TestLiveCountsConcurrentReadWrite runs writers and stats readers at once so
 // the race detector proves the counters are safe under concurrent use.
 func TestLiveCountsConcurrentReadWrite(t *testing.T) {
+	t.Parallel()
+
 	cfg := &config.Config{StateDir: t.TempDir()}
 
 	db, err := New(cfg, logger.New())
@@ -412,17 +727,21 @@ func TestLiveCountsConcurrentReadWrite(t *testing.T) {
 	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
 
 	const writers = 4
-	var wg sync.WaitGroup
 
+	var wg sync.WaitGroup
 	wg.Add(writers)
+
 	for w := range writers {
 		go func(base int) {
 			defer wg.Done()
+
 			for i := range 25 {
 				asn := 65000 + base*100 + i
-				route := mkV6Route("2001:db8::/32", asn, ts)
+				route := mkV6Route(testV6Prefix, asn, ts)
 				route.PeerIP = "2001:db8::" + uuid.NewString()
-				if err := db.UpsertLiveRoute(route); err != nil {
+
+				err := db.UpsertLiveRoute(route)
+				if err != nil {
 					t.Errorf("UpsertLiveRoute: %v", err)
 
 					return
@@ -433,15 +752,19 @@ func TestLiveCountsConcurrentReadWrite(t *testing.T) {
 
 	var readerWG sync.WaitGroup
 	readerWG.Add(1)
+
 	stop := make(chan struct{})
+
 	go func() {
 		defer readerWG.Done()
+
 		for {
 			select {
 			case <-stop:
 				return
 			default:
-				if _, err := db.GetStatsContext(context.Background()); err != nil {
+				_, err := db.GetStatsContext(context.Background())
+				if err != nil {
 					t.Errorf("GetStatsContext: %v", err)
 
 					return
@@ -458,6 +781,7 @@ func TestLiveCountsConcurrentReadWrite(t *testing.T) {
 	if err != nil {
 		t.Fatalf("GetStatsContext: %v", err)
 	}
+
 	if want := writers * 25; stats.IPv6Routes != want {
 		t.Errorf("IPv6Routes = %d, want %d", stats.IPv6Routes, want)
 	}
@@ -482,21 +806,27 @@ func assertCounts(t *testing.T, when string, got Stats, want wantCounts) {
 	if got.ASNs != want.asns {
 		t.Errorf("%s: ASNs = %d, want %d", when, got.ASNs, want.asns)
 	}
+
 	if got.Prefixes != want.prefixes {
 		t.Errorf("%s: Prefixes = %d, want %d", when, got.Prefixes, want.prefixes)
 	}
+
 	if got.Peerings != want.peerings {
 		t.Errorf("%s: Peerings = %d, want %d", when, got.Peerings, want.peerings)
 	}
+
 	if got.Peers != want.peers {
 		t.Errorf("%s: Peers = %d, want %d", when, got.Peers, want.peers)
 	}
+
 	if got.IPv4Routes != want.ipv4Routes {
 		t.Errorf("%s: IPv4Routes = %d, want %d", when, got.IPv4Routes, want.ipv4Routes)
 	}
+
 	if got.IPv6Routes != want.ipv6Routes {
 		t.Errorf("%s: IPv6Routes = %d, want %d", when, got.IPv6Routes, want.ipv6Routes)
 	}
+
 	if got.LiveRoutes != want.liveRoutes {
 		t.Errorf("%s: LiveRoutes = %d, want %d", when, got.LiveRoutes, want.liveRoutes)
 	}

internal/database/database_test.go

@@ -1,319 +0,0 @@
-package database
-
-import (
-	"context"
-	"database/sql"
-	"errors"
-	"fmt"
-	"net/netip"
-	"sync"
-	"testing"
-	"time"
-
-	"git.eeqj.de/sneak/routewatch/internal/config"
-	"git.eeqj.de/sneak/routewatch/internal/logger"
-	"github.com/google/uuid"
-)
-
-// tempStoreMemory is the PRAGMA temp_store value meaning "hold temp B-trees in
-// memory"; the DSN change must leave temp_store below this so they spill to disk.
-const tempStoreMemory = 2
-
-// autoVacuumIncremental is the PRAGMA auto_vacuum value meaning "incremental".
-const autoVacuumIncremental = 2
-
-// vacuumTestRoutes is how many routes the vacuum test writes and then deletes,
-// enough to leave many free pages in the file.
-const vacuumTestRoutes = 2000
-
-// heldConnections is how many pooled connections the pragma test holds open at
-// once so each is a distinct SQLite connection that parsed the DSN.
-const heldConnections = 5
-
-// testPeerIP is the peer every route in the IP lookup test is learned from.
-const testPeerIP = "192.0.2.254"
-
-// Parameters for the checkpoint-contention regression test.
-const (
-	// contentionIterations is how many batch writes race the checkpoint loop.
-	contentionIterations = 400
-	// contendedASNCount is the small set of ASNs the batches reuse, so most
-	// batches update existing rows and exercise the read-before-write path.
-	contendedASNCount = 16
-	// asnSecondBand offsets a second ASN per batch so each batch writes more
-	// than one row.
-	asnSecondBand = 100
-)
-
-// TestGetIPInfoFindsMostSpecificLiveRoute stores nested live prefixes for both
-// families and checks that a lookup returns the most specific one covering the
-// address, ErrNoRoute when none covers it, and the next less specific prefix
-// once the only route of the most specific one is withdrawn.
-func TestGetIPInfoFindsMostSpecificLiveRoute(t *testing.T) {
-	cfg := &config.Config{StateDir: t.TempDir()}
-
-	db, err := New(cfg, logger.New())
-	if err != nil {
-		t.Fatalf("failed to create database: %v", err)
-	}
-	defer func() { _ = db.Close() }()
-
-	// Nested live prefixes, each originated by its own AS.
-	origins := map[string]int{
-		"10.0.0.0/8":        64500,
-		"10.1.0.0/16":       64501,
-		"10.1.2.0/24":       64502,
-		"2001:db8::/32":     64500,
-		"2001:db8:1::/48":   64501,
-		"2001:db8:1:2::/64": 64502,
-	}
-	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
-	routes := make([]*LiveRoute, 0, len(origins))
-	for prefix, asn := range origins {
-		routes = append(routes, &LiveRoute{
-			ID:          uuid.New(),
-			Prefix:      prefix,
-			MaskLength:  netip.MustParsePrefix(prefix).Bits(),
-			IPVersion:   detectIPVersion(prefix),
-			OriginASN:   asn,
-			PeerIP:      testPeerIP,
-			ASPath:      []int{asn},
-			NextHop:     testPeerIP,
-			LastUpdated: ts,
-		})
-	}
-	if err := db.UpsertLiveRouteBatch(routes); err != nil {
-		t.Fatalf("UpsertLiveRouteBatch: %v", err)
-	}
-
-	// lookup checks that ip resolves to the live prefix want, or to ErrNoRoute
-	// when want is empty.
-	lookup := func(ip, want string) {
-		t.Helper()
-
-		info, err := db.GetIPInfo(ip)
-		if want == "" {
-			if !errors.Is(err, ErrNoRoute) {
-				t.Errorf("GetIPInfo(%s) = %+v, %v; want ErrNoRoute", ip, info, err)
-			}
-
-			return
-		}
-		if err != nil {
-			t.Errorf("GetIPInfo(%s): %v", ip, err)
-
-			return
-		}
-		if info.Netblock != want || info.MaskLength != netip.MustParsePrefix(want).Bits() ||
-			info.ASN != origins[want] {
-			t.Errorf("GetIPInfo(%s) = %s (mask %d) AS%d, want %s AS%d",
-				ip, info.Netblock, info.MaskLength, info.ASN, want, origins[want])
-		}
-	}
-
-	lookup("10.1.2.3", "10.1.2.0/24")
-	lookup("::ffff:10.1.2.3", "10.1.2.0/24")
-	lookup("10.1.3.4", "10.1.0.0/16")
-	lookup("10.2.0.1", "10.0.0.0/8")
-	lookup("192.0.2.1", "")
-	lookup("2001:db8:1:2::3", "2001:db8:1:2::/64")
-	lookup("2001:db8:1:3::4", "2001:db8:1::/48")
-	lookup("2001:db8:2::1", "2001:db8::/32")
-	lookup("2001:db9::1", "")
-
-	err = db.DeleteLiveRouteBatch([]LiveRouteDeletion{
-		{Prefix: "10.1.2.0/24", OriginASN: 64502, PeerIP: testPeerIP, IPVersion: ipVersionV4},
-		{Prefix: "2001:db8:1:2::/64", OriginASN: 64502, PeerIP: testPeerIP, IPVersion: ipVersionV6},
-	})
-	if err != nil {
-		t.Fatalf("DeleteLiveRouteBatch: %v", err)
-	}
-
-	lookup("10.1.2.3", "10.1.0.0/16")
-	lookup("2001:db8:1:2::3", "2001:db8:1::/48")
-}
-
-// TestConnectionPoolPragmas holds several pooled connections open at once and
-// checks each one carries the per-connection settings from the DSN, plus the
-// process-wide hard heap limit, and sees the new file with auto_vacuum
-// incremental.
-func TestConnectionPoolPragmas(t *testing.T) {
-	cfg := &config.Config{StateDir: t.TempDir()}
-
-	db, err := New(cfg, logger.New())
-	if err != nil {
-		t.Fatalf("failed to create database: %v", err)
-	}
-	defer func() { _ = db.Close() }()
-
-	ctx := context.Background()
-
-	// Hold distinct connections open simultaneously so the pool must open a new
-	// one (each parsing the DSN) rather than hand back the same connection.
-	conns := make([]*sql.Conn, 0, heldConnections)
-	defer func() {
-		for _, c := range conns {
-			_ = c.Close()
-		}
-	}()
-
-	for i := 0; i < heldConnections; i++ {
-		c, err := db.db.Conn(ctx)
-		if err != nil {
-			t.Fatalf("failed to open connection %d: %v", i, err)
-		}
-		conns = append(conns, c)
-	}
-
-	for i, c := range conns {
-		var cacheSize int
-		if err := c.QueryRowContext(ctx, "PRAGMA cache_size").Scan(&cacheSize); err != nil {
-			t.Fatalf("conn %d: failed to read cache_size: %v", i, err)
-		}
-		if cacheSize != sqliteCacheSizeKiB {
-			t.Errorf("conn %d: cache_size = %d, want %d", i, cacheSize, sqliteCacheSizeKiB)
-		}
-
-		var busyTimeout int
-		if err := c.QueryRowContext(ctx, "PRAGMA busy_timeout").Scan(&busyTimeout); err != nil {
-			t.Fatalf("conn %d: failed to read busy_timeout: %v", i, err)
-		}
-		if busyTimeout != sqliteBusyTimeoutMs {
-			t.Errorf("conn %d: busy_timeout = %d, want %d", i, busyTimeout, sqliteBusyTimeoutMs)
-		}
-
-		var tempStore int
-		if err := c.QueryRowContext(ctx, "PRAGMA temp_store").Scan(&tempStore); err != nil {
-			t.Fatalf("conn %d: failed to read temp_store: %v", i, err)
-		}
-		if tempStore == tempStoreMemory {
-			t.Errorf("conn %d: temp_store = %d, want anything but %d (MEMORY)", i, tempStore, tempStoreMemory)
-		}
-
-		var hardHeapLimit int64
-		if err := c.QueryRowContext(ctx, "PRAGMA hard_heap_limit").Scan(&hardHeapLimit); err != nil {
-			t.Fatalf("conn %d: failed to read hard_heap_limit: %v", i, err)
-		}
-		if hardHeapLimit != sqliteHardHeapLimitBytes {
-			t.Errorf("conn %d: hard_heap_limit = %d, want %d", i, hardHeapLimit, sqliteHardHeapLimitBytes)
-		}
-
-		var autoVacuum int
-		if err := c.QueryRowContext(ctx, "PRAGMA auto_vacuum").Scan(&autoVacuum); err != nil {
-			t.Fatalf("conn %d: failed to read auto_vacuum: %v", i, err)
-		}
-		if autoVacuum != autoVacuumIncremental {
-			t.Errorf("conn %d: auto_vacuum = %d, want %d (incremental)", i, autoVacuum, autoVacuumIncremental)
-		}
-	}
-}
-
-// TestVacuumReturnsFreePages checks that after routes are deleted from a new
-// database, one Vacuum call returns every page they used. The deletes leave
-// fewer free pages than the 1000 Vacuum frees per call, so none may remain.
-// With auto_vacuum off (issue https://git.eeqj.de/sneak/routewatch/issues/43)
-// the free pages stayed in the file, and with the PRAGMA run by ExecContext
-// Vacuum freed only one page per call.
-func TestVacuumReturnsFreePages(t *testing.T) {
-	cfg := &config.Config{StateDir: t.TempDir()}
-
-	db, err := New(cfg, logger.New())
-	if err != nil {
-		t.Fatalf("failed to create database: %v", err)
-	}
-	defer func() { _ = db.Close() }()
-
-	ctx := context.Background()
-	ts := time.Now().UTC()
-	const asn = 64500
-
-	routes := make([]*LiveRoute, 0, vacuumTestRoutes)
-	deletions := make([]LiveRouteDeletion, 0, vacuumTestRoutes)
-	for i := range vacuumTestRoutes {
-		route := mkV4Route(t, fmt.Sprintf("10.%d.%d.0/24", i/256, i%256), asn, ts)
-		routes = append(routes, route)
-		deletions = append(deletions, LiveRouteDeletion{
-			Prefix:    route.Prefix,
-			OriginASN: asn,
-			PeerIP:    route.PeerIP,
-			IPVersion: ipVersionV4,
-		})
-	}
-
-	if err := db.UpsertLiveRouteBatch(routes); err != nil {
-		t.Fatalf("failed to write routes: %v", err)
-	}
-	if err := db.DeleteLiveRouteBatch(deletions); err != nil {
-		t.Fatalf("failed to delete routes: %v", err)
-	}
-
-	var before int
-	if err := db.db.QueryRowContext(ctx, "PRAGMA freelist_count").Scan(&before); err != nil {
-		t.Fatalf("failed to read freelist_count: %v", err)
-	}
-	if before == 0 {
-		t.Fatalf("no free pages after deleting %d routes", vacuumTestRoutes)
-	}
-
-	if err := db.Vacuum(ctx); err != nil {
-		t.Fatalf("Vacuum failed: %v", err)
-	}
-
-	var after int
-	if err := db.db.QueryRowContext(ctx, "PRAGMA freelist_count").Scan(&after); err != nil {
-		t.Fatalf("failed to read freelist_count: %v", err)
-	}
-	if after != 0 {
-		t.Errorf("free pages after Vacuum = %d of %d, want 0", after, before)
-	}
-}
-
-// TestBatchWriteDuringCheckpoint reproduces issue #25. A batch write reads
-// (SELECT) before it writes (INSERT/UPDATE). Under the default deferred locking
-// the transaction begins as a reader and, when the maintainer's WAL checkpoint
-// holds the write lock, its upgrade to writer fails immediately with "database
-// is locked" without honouring busy_timeout, dropping the batch. With
-// _txlock=immediate the transaction takes the write lock at BEGIN and waits, so
-// no batch is dropped. The checkpoint runs without the Database mutex, exactly
-// as the background maintainer does in production.
-func TestBatchWriteDuringCheckpoint(t *testing.T) {
-	cfg := &config.Config{StateDir: t.TempDir()}
-
-	db, err := New(cfg, logger.New())
-	if err != nil {
-		t.Fatalf("failed to create database: %v", err)
-	}
-	defer func() { _ = db.Close() }()
-
-	ctx, cancel := context.WithCancel(context.Background())
-
-	var wg sync.WaitGroup
-	wg.Add(1)
-	go func() {
-		defer wg.Done()
-		for {
-			select {
-			case <-ctx.Done():
-				return
-			default:
-				_ = db.Checkpoint(ctx) // errors are the checkpoint's own to absorb
-			}
-		}
-	}()
-
-	ts := time.Now().UTC()
-	for i := 0; i < contentionIterations; i++ {
-		asns := map[int]time.Time{
-			i % contendedASNCount:                   ts,
-			(i % contendedASNCount) + asnSecondBand: ts,
-		}
-		if err := db.GetOrCreateASNBatch(asns); err != nil {
-			cancel()
-			wg.Wait()
-			t.Fatalf("batch write failed under checkpoint contention: %v", err)
-		}
-	}
-
-	cancel()
-	wg.Wait()
-}

internal/database/dockerignore_internal_test.go

@@ -0,0 +1,61 @@
+package database
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/config"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+)
+
+// TestDockerignoreExcludesDatabaseFiles checks that .dockerignore keeps the
+// database file, and the WAL and shared-memory files SQLite writes next to it,
+// out of the build context wherever they sit in the clone. The image ships the
+// whole build context as a source archive.
+func TestDockerignoreExcludesDatabaseFiles(t *testing.T) {
+	t.Parallel()
+
+	db, err := New(&config.Config{StateDir: t.TempDir()}, logger.New())
+	if err != nil {
+		t.Fatalf("failed to create database: %v", err)
+	}
+	defer func() { _ = db.Close() }()
+
+	data, err := os.ReadFile("../../.dockerignore")
+	if err != nil {
+		t.Fatalf("failed to read .dockerignore: %v", err)
+	}
+
+	// Docker reads `**/` as any number of directories and the rest of the
+	// pattern as filepath.Match does, so `**/` followed by a pattern that
+	// matches a file's name excludes that file at any depth.
+	var anyDepth []string
+
+	for line := range strings.Lines(string(data)) {
+		pattern, ok := strings.CutPrefix(strings.TrimSpace(line), "**/")
+		if ok {
+			anyDepth = append(anyDepth, pattern)
+		}
+	}
+
+	name := filepath.Base(db.path)
+
+	for _, file := range []string{name, name + "-wal", name + "-shm"} {
+		excluded := false
+
+		for _, pattern := range anyDepth {
+			matched, err := filepath.Match(pattern, file)
+			if err != nil {
+				t.Fatalf("bad .dockerignore pattern **/%s: %v", pattern, err)
+			}
+
+			excluded = excluded || matched
+		}
+
+		if !excluded {
+			t.Errorf(".dockerignore has no **/ pattern that excludes %s", file)
+		}
+	}
+}

internal/database/interface.go

@@ -1,10 +1,14 @@
 // Package database provides SQLite storage for BGP routing data including ASNs,
-// prefixes, announcements, peerings, and live route tables.
+// prefixes, announcements, peerings, and live route tables. Its routes table
+// holds a copy of the whole routing table that internal/routetable keeps in
+// memory.
 package database
 
 import (
 	"context"
 	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
 )
 
 // Stats contains database statistics including counts for ASNs, prefixes,
@@ -29,10 +33,12 @@ type Stats struct {
 // Store defines the interface for database operations. It provides methods for
 // managing ASNs, prefixes, announcements, peerings, BGP peers, and live routes.
 // Implementations must be safe for concurrent use.
+//
+//nolint:interfacebloat // the one seam handlers and tests depend on
 type Store interface {
 	// ASN operations
 	GetOrCreateASN(number int, timestamp time.Time) (*ASN, error)
-	GetOrCreateASNBatch(asns map[int]time.Time) error
+	GetOrCreateASNBatch(asns map[int]SeenTimes) error
 
 	// Prefix operations
 	GetOrCreatePrefix(prefix string, timestamp time.Time) (*Prefix, error)
@@ -42,7 +48,7 @@ type Store interface {
 	RecordAnnouncement(announcement *Announcement) error
 
 	// Peering operations
-	RecordPeering(asA, asB int, timestamp time.Time) error
+	RecordPeering(asA, asB int, seen SeenTimes) error
 
 	// Statistics
 	GetStats() (Stats, error)
@@ -55,13 +61,20 @@ type Store interface {
 	// Live route operations
 	UpsertLiveRoute(route *LiveRoute) error
 	UpsertLiveRouteBatch(routes []*LiveRoute) error
-	DeleteLiveRoute(prefix string, originASN int, peerIP string) error
+	DeleteLiveRoute(prefix, peerIP string) error
 	DeleteLiveRouteBatch(deletions []LiveRouteDeletion) error
-	GetPrefixDistribution() (ipv4 []PrefixDistribution, ipv6 []PrefixDistribution, err error)
+	DeleteLiveRoutesOfPeer(peerIP string) (deleted int, err error)
-	GetPrefixDistributionContext(ctx context.Context) (ipv4 []PrefixDistribution, ipv6 []PrefixDistribution, err error)
+	GetPrefixDistribution() (
+		ipv4 []PrefixDistribution, ipv6 []PrefixDistribution, err error)
+	GetPrefixDistributionContext(ctx context.Context) (
+		ipv4 []PrefixDistribution, ipv6 []PrefixDistribution, err error)
 	GetLiveRouteCounts() (ipv4Count, ipv6Count int, err error)
 	GetLiveRouteCountsContext(ctx context.Context) (ipv4Count, ipv6Count int, err error)
 
+	// Routing table operations: the database's copy of the table in memory
+	WriteRouteChanges(changes []routetable.Change, newPeers map[string]uint) error
+	ReadRouteTable(ctx context.Context) (*routetable.Table, error)
+
 	// IP lookup operations
 	GetIPInfo(ip string) (*IPInfo, error)
 	GetIPInfoContext(ctx context.Context, ip string) (*IPInfo, error)
@@ -76,10 +89,14 @@ type Store interface {
 	GetASDetailsContext(ctx context.Context, asn int) (*ASN, []LiveRoute, error)
 	GetASPeers(asn int) ([]ASPeer, error)
 	GetASPeersContext(ctx context.Context, asn int) ([]ASPeer, error)
-	GetPrefixDetails(prefix string) ([]LiveRoute, error)
+	GetPrefixDetails(prefix string) ([]LiveRoute, map[int]ASN, error)
-	GetPrefixDetailsContext(ctx context.Context, prefix string) ([]LiveRoute, error)
+	GetPrefixDetailsContext(
+		ctx context.Context, prefix string,
+	) ([]LiveRoute, map[int]ASN, error)
 	GetRandomPrefixesByLength(maskLength, ipVersion, limit int) ([]LiveRoute, error)
-	GetRandomPrefixesByLengthContext(ctx context.Context, maskLength, ipVersion, limit int) ([]LiveRoute, error)
+	GetRandomPrefixesByLengthContext(
+		ctx context.Context, maskLength, ipVersion, limit int,
+	) ([]LiveRoute, error)
 
 	// Lifecycle
 	Close() error

internal/database/models.go

@@ -1,4 +1,3 @@
-// Package database provides SQLite storage for BGP routing data.
 package database
 
 import (
@@ -9,6 +8,8 @@ import (
 
 // ASN represents an Autonomous System Number with its metadata including
 // handle, description, WHOIS data, and first/last seen timestamps.
+//
+//nolint:tagliatelle // field names are the existing JSON API
 type ASN struct {
 	ASN         int    `json:"asn"`
 	Handle      string `json:"handle"`
@@ -27,48 +28,48 @@ type ASN struct {
 	RIRRegDate  *time.Time `json:"rir_registration_date,omitempty"`
 	RIRLastMod  *time.Time `json:"rir_last_modified,omitempty"`
 	WHOISRaw    string     `json:"whois_raw,omitempty"`
-	// Timestamps
+	// Timestamps. They are zero, and left out of JSON, for an AS that has no
-	FirstSeen      time.Time  `json:"first_seen"`
+	// record in the asns table.
-	LastSeen       time.Time  `json:"last_seen"`
+	FirstSeen      time.Time  `json:"first_seen,omitzero"`
+	LastSeen       time.Time  `json:"last_seen,omitzero"`
 	WHOISUpdatedAt *time.Time `json:"whois_updated_at,omitempty"`
 }
 
 // Prefix represents an IP prefix (CIDR block) with its IP version (4 or 6)
-// and first/last seen timestamps.
+// and the time it was first seen.
 type Prefix struct {
-	ID        uuid.UUID `json:"id"`
+	Prefix    string
-	Prefix    string    `json:"prefix"`
+	IPVersion int // 4 or 6
-	IPVersion int       `json:"ip_version"` // 4 or 6
+	FirstSeen time.Time
-	FirstSeen time.Time `json:"first_seen"`
-	LastSeen  time.Time `json:"last_seen"`
 }
 
 // Announcement represents a BGP announcement or withdrawal event,
 // containing the prefix, AS path, origin ASN, peer ASN, next hop, and timestamp.
 type Announcement struct {
-	ID           uuid.UUID `json:"id"`
+	ID           uuid.UUID
-	PrefixID     uuid.UUID `json:"prefix_id"`
+	PrefixID     uuid.UUID
-	PeerASN      int       `json:"peer_asn"`
+	PeerASN      int
-	OriginASN    int       `json:"origin_asn"`
+	OriginASN    int
-	Path         string    `json:"path"` // JSON-encoded AS path
+	Path         string // JSON-encoded AS path
-	NextHop      string    `json:"next_hop"`
+	NextHop      string
-	Timestamp    time.Time `json:"timestamp"`
+	Timestamp    time.Time
-	IsWithdrawal bool      `json:"is_withdrawal"`
+	IsWithdrawal bool
 }
 
 // ASNPeering represents a peering relationship between two ASNs,
 // stored with the lower ASN as ASA and the higher as ASB.
 type ASNPeering struct {
-	ID        uuid.UUID `json:"id"`
+	ID        uuid.UUID
-	ASA       int       `json:"as_a"`
+	ASA       int
-	ASB       int       `json:"as_b"`
+	ASB       int
-	FirstSeen time.Time `json:"first_seen"`
+	FirstSeen time.Time
-	LastSeen  time.Time `json:"last_seen"`
+	LastSeen  time.Time
 }
 
 // LiveRoute represents a route in the live routing table
+//
+//nolint:tagliatelle // field names are the existing JSON API
 type LiveRoute struct {
-	ID          uuid.UUID `json:"id"`
 	Prefix      string    `json:"prefix"`
 	MaskLength  int       `json:"mask_length"`
 	IPVersion   int       `json:"ip_version"`
@@ -80,15 +81,18 @@ type LiveRoute struct {
 }
 
 // PrefixDistribution represents the distribution of prefixes by mask length
+//
+//nolint:tagliatelle // field names are the existing JSON API
 type PrefixDistribution struct {
 	MaskLength int `json:"mask_length"`
 	Count      int `json:"count"`
 }
 
 // IPInfo represents comprehensive IP information for the /ip endpoint
+//
+//nolint:tagliatelle // field names are the existing JSON API
 type IPInfo struct {
 	IP         string `json:"ip"`
-	PTR        []string `json:"ptr,omitempty"`
 	Netblock   string `json:"netblock"`
 	MaskLength int    `json:"mask_length"`
 	IPVersion  int    `json:"ip_version"`
@@ -114,17 +118,39 @@ type IPInfo struct {
 // LiveRouteDeletion represents parameters for deleting a live route
 type LiveRouteDeletion struct {
 	Prefix    string
-	OriginASN int
 	PeerIP    string
 	IPVersion int
 }
 
-// PeerUpdate represents parameters for updating a peer
+// SeenTimes is the earliest and the latest time one batch of updates saw an AS,
+// a peer or a peering. A row the batch adds is stored first seen at First and
+// last seen at Last. A row already stored takes First as its first_seen only
+// when First is earlier, and Last as its last_seen only when Last is later.
+type SeenTimes struct {
+	First time.Time
+	Last  time.Time
+}
+
+// Add makes t the earliest time of s when it is earlier than First, and the
+// latest when it is later than Last. The zero SeenTimes holds no time yet, so
+// the first Add sets both.
+func (s *SeenTimes) Add(t time.Time) {
+	if s.First.IsZero() || t.Before(s.First) {
+		s.First = t
+	}
+
+	if s.Last.IsZero() || t.After(s.Last) {
+		s.Last = t
+	}
+}
+
+// PeerUpdate represents parameters for updating a peer. PeerASN and
+// MessageType are those of the peer's latest update.
 type PeerUpdate struct {
 	PeerIP      string
 	PeerASN     int
 	MessageType string
-	Timestamp   time.Time
+	Seen        SeenTimes
 }
 
 // ASNWHOISUpdate contains WHOIS data for updating an ASN record.

internal/database/routes.go

@@ -0,0 +1,339 @@
+package database
+
+import (
+	"context"
+	"database/sql"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"net/netip"
+	"time"
+
+	"github.com/bits-and-blooms/bitset"
+
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+// keptRoutesPerRow is how many kept routes a row of the routes table has
+// columns for: the newest and the older one.
+const keptRoutesPerRow = 2
+
+// errUnknownRoutePeer is returned for a peer that route_peers lacks: on a
+// write, the peer of a kept route; on a read, a peer number in a row.
+var errUnknownRoutePeer = errors.New("peer not in route_peers")
+
+// The statements of the routes and route_peers tables (schema.sql).
+const (
+	insertRoutePeerSQL = `INSERT INTO route_peers (number, address)
+		VALUES (?, ?)`
+	routePeerNumberSQL = `SELECT number FROM route_peers WHERE address = ?`
+	routePeersSQL      = `SELECT number, address FROM route_peers`
+
+	upsertRouteSQL = `INSERT INTO routes (prefix, origin_asn, peers,
+			newest_peer, newest_as_path, newest_time,
+			older_peer, older_as_path, older_time)
+		VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
+		ON CONFLICT (prefix, origin_asn) DO UPDATE SET peers = excluded.peers,
+			newest_peer = excluded.newest_peer,
+			newest_as_path = excluded.newest_as_path,
+			newest_time = excluded.newest_time,
+			older_peer = excluded.older_peer,
+			older_as_path = excluded.older_as_path,
+			older_time = excluded.older_time`
+	deleteRouteSQL = `DELETE FROM routes WHERE prefix = ? AND origin_asn = ?`
+	routesSQL      = `SELECT prefix, origin_asn, peers,
+		newest_peer, newest_as_path, newest_time,
+		older_peer, older_as_path, older_time FROM routes`
+)
+
+// WriteRouteChanges writes, in one transaction, what one call of the routing
+// table's TakeChanges returned. It stores each new peer with its number, writes
+// the row of each changed prefix and origin AS pair, and deletes the row of
+// each removed one. Route times are stored to the second, as the feed gives
+// them.
+//
+// When it fails, it writes nothing. The caller then writes these changes and
+// peers again, because later changes can name the peers it did not store.
+func (d *Database) WriteRouteChanges(
+	changes []routetable.Change, newPeers map[string]uint,
+) error {
+	if len(changes) == 0 && len(newPeers) == 0 {
+		return nil
+	}
+
+	d.lock("WriteRouteChanges")
+	defer d.unlock()
+
+	tx, err := d.beginTx()
+	if err != nil {
+		return fmt.Errorf("failed to begin transaction: %w", err)
+	}
+	defer d.rollback(tx)
+
+	ctx := context.Background()
+
+	// The new peers first, so that the rows can look up their numbers.
+	for address, number := range newPeers {
+		_, err := tx.ExecContext(ctx, insertRoutePeerSQL, number, address)
+		if err != nil {
+			return fmt.Errorf("failed to store peer %s: %w", address, err)
+		}
+	}
+
+	err = writeRouteRows(ctx, tx, changes)
+	if err != nil {
+		return err
+	}
+
+	err = tx.Commit()
+	if err != nil {
+		return fmt.Errorf("failed to commit transaction: %w", err)
+	}
+
+	return nil
+}
+
+// writeRouteRows upserts the row of each change in tx, or deletes it for a
+// removed pair.
+func writeRouteRows(
+	ctx context.Context, tx *loggingTx, changes []routetable.Change,
+) error {
+	upsert, err := tx.PrepareContext(ctx, upsertRouteSQL)
+	if err != nil {
+		return fmt.Errorf("failed to prepare route upsert: %w", err)
+	}
+	defer func() { _ = upsert.Close() }()
+
+	del, err := tx.PrepareContext(ctx, deleteRouteSQL)
+	if err != nil {
+		return fmt.Errorf("failed to prepare route delete: %w", err)
+	}
+	defer func() { _ = del.Close() }()
+
+	peerNumber, err := tx.PrepareContext(ctx, routePeerNumberSQL)
+	if err != nil {
+		return fmt.Errorf("failed to prepare peer number lookup: %w", err)
+	}
+	defer func() { _ = peerNumber.Close() }()
+
+	for _, change := range changes {
+		prefix := change.Prefix.String()
+
+		if change.Removed {
+			_, err := del.ExecContext(ctx, prefix, change.OriginASN)
+			if err != nil {
+				return fmt.Errorf("failed to delete route %s: %w", prefix, err)
+			}
+
+			continue
+		}
+
+		row, err := routeRow(ctx, peerNumber, change)
+		if err != nil {
+			return fmt.Errorf("failed to write route %s: %w", prefix, err)
+		}
+
+		_, err = upsert.ExecContext(ctx, row...)
+		if err != nil {
+			return fmt.Errorf("failed to write route %s: %w", prefix, err)
+		}
+	}
+
+	return nil
+}
+
+// routeRow returns the values of the routes row of change, in the order of
+// upsertRouteSQL. peerNumber looks up a peer's number by its address. The
+// table keeps at most keptRoutesPerRow routes of a pair.
+func routeRow(
+	ctx context.Context, peerNumber *sql.Stmt, change routetable.Change,
+) ([]any, error) {
+	peers, err := change.Peers.MarshalBinary()
+	if err != nil {
+		return nil, err
+	}
+
+	row := []any{change.Prefix.String(), change.OriginASN, peers}
+
+	for i := range keptRoutesPerRow {
+		if i >= len(change.Routes) {
+			row = append(row, nil, nil, nil)
+
+			continue
+		}
+
+		r := change.Routes[i]
+
+		var number uint
+
+		err := peerNumber.QueryRowContext(ctx, r.Peer).Scan(&number)
+		if errors.Is(err, sql.ErrNoRows) {
+			return nil, fmt.Errorf("%w: %s", errUnknownRoutePeer, r.Peer)
+		}
+
+		if err != nil {
+			return nil, err
+		}
+
+		path, err := json.Marshal(r.ASPath)
+		if err != nil {
+			return nil, err
+		}
+
+		row = append(row, number, string(path), r.Time.Unix())
+	}
+
+	return row, nil
+}
+
+// ReadRouteTable returns the routing table that the routes and route_peers
+// tables hold, as startup reads it. It reads the peers first, so that the
+// table keeps their numbers, then loads each row into the table as it reads
+// it. The two reads are separate, so nothing may write the tables meanwhile.
+// It fails on a row that names a peer number route_peers lacks.
+func (d *Database) ReadRouteTable(ctx context.Context) (*routetable.Table, error) {
+	peerNumbers, err := d.readRoutePeers(ctx)
+	if err != nil {
+		return nil, err
+	}
+
+	addresses := make(map[uint]string, len(peerNumbers))
+	numbers := new(bitset.BitSet)
+
+	for address, number := range peerNumbers {
+		addresses[number] = address
+		numbers.Set(number)
+	}
+
+	table := routetable.New(routetable.Params{PeerNumbers: peerNumbers})
+
+	rows, err := d.db.QueryContext(ctx, routesSQL)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read routes: %w", err)
+	}
+	defer func() { _ = rows.Close() }()
+
+	for rows.Next() {
+		change, err := scanRouteRow(rows, addresses, numbers)
+		if err != nil {
+			return nil, err
+		}
+
+		table.Load(change)
+	}
+
+	err = rows.Err()
+	if err != nil {
+		return nil, fmt.Errorf("failed to read routes: %w", err)
+	}
+
+	return table, nil
+}
+
+// readRoutePeers returns the number of each peer in route_peers, by address.
+func (d *Database) readRoutePeers(ctx context.Context) (map[string]uint, error) {
+	rows, err := d.db.QueryContext(ctx, routePeersSQL)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read route peers: %w", err)
+	}
+	defer func() { _ = rows.Close() }()
+
+	peerNumbers := make(map[string]uint)
+
+	for rows.Next() {
+		var (
+			number  uint
+			address string
+		)
+
+		err := rows.Scan(&number, &address)
+		if err != nil {
+			return nil, fmt.Errorf("failed to read a route peer: %w", err)
+		}
+
+		peerNumbers[address] = number
+	}
+
+	err = rows.Err()
+	if err != nil {
+		return nil, fmt.Errorf("failed to read route peers: %w", err)
+	}
+
+	return peerNumbers, nil
+}
+
+// keptRouteColumns are the three columns of one kept route in a routes row.
+type keptRouteColumns struct {
+	peer   sql.Null[uint]
+	asPath sql.NullString
+	time   sql.NullInt64
+}
+
+// scanRouteRow returns the content of the routes row rows is at, with each kept
+// route's peer named by its address in addresses. It fails when the row names a
+// peer number that addresses lacks. numbers holds the numbers of addresses as a
+// set, against which a row's whole peer set is checked at once.
+func scanRouteRow(
+	rows *sql.Rows, addresses map[uint]string, numbers *bitset.BitSet,
+) (routetable.Change, error) {
+	var (
+		change routetable.Change
+		prefix string
+		peers  []byte
+		kept   [keptRoutesPerRow]keptRouteColumns
+	)
+
+	err := rows.Scan(&prefix, &change.OriginASN, &peers,
+		&kept[0].peer, &kept[0].asPath, &kept[0].time,
+		&kept[1].peer, &kept[1].asPath, &kept[1].time)
+	if err != nil {
+		return change, fmt.Errorf("failed to read a route: %w", err)
+	}
+
+	change.Prefix, err = netip.ParsePrefix(prefix)
+	if err != nil {
+		return change, fmt.Errorf("failed to read route %s: %w", prefix, err)
+	}
+
+	change.Peers = new(bitset.BitSet)
+
+	err = change.Peers.UnmarshalBinary(peers)
+	if err != nil {
+		return change, fmt.Errorf("failed to read the peers of route %s: %w",
+			prefix, err)
+	}
+
+	if !numbers.IsSuperSet(change.Peers) {
+		unknown, _ := change.Peers.Difference(numbers).NextSet(0)
+
+		return change, fmt.Errorf("%w: number %d in the peers of route %s",
+			errUnknownRoutePeer, unknown, prefix)
+	}
+
+	for _, columns := range kept {
+		if !columns.peer.Valid {
+			continue
+		}
+
+		address, known := addresses[columns.peer.V]
+		if !known {
+			return change, fmt.Errorf("%w: number %d in route %s",
+				errUnknownRoutePeer, columns.peer.V, prefix)
+		}
+
+		r := routetable.Route{
+			Peer: address,
+			Time: time.Unix(columns.time.Int64, 0).UTC(),
+		}
+
+		err := json.Unmarshal([]byte(columns.asPath.String), &r.ASPath)
+		if err != nil {
+			return change, fmt.Errorf("failed to read an AS path of route %s: %w",
+				prefix, err)
+		}
+
+		change.Routes = append(change.Routes, r)
+	}
+
+	return change, nil
+}

internal/database/routes_internal_test.go

@@ -0,0 +1,527 @@
+package database
+
+import (
+	"errors"
+	"fmt"
+	"math"
+	"math/rand/v2"
+	"net/netip"
+	"os"
+	"reflect"
+	"slices"
+	"testing"
+	"time"
+
+	"github.com/bits-and-blooms/bitset"
+
+	"git.eeqj.de/sneak/routewatch/internal/config"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+// Peers, origin ASes and prefixes of the routes table tests.
+const (
+	peerA = "192.0.2.1"
+	peerB = "192.0.2.2"
+	peerC = "2001:db8::3"
+	peerD = "2001:db8::4"
+	peerE = "2001:db8::5"
+
+	asA = 64500
+	asB = 64501
+	// transitAS is the first AS of every test path, the peer's own.
+	transitAS = 64496
+
+	tablePrefixV4      = "10.1.2.0/24"
+	tablePrefixShortV4 = "10.9.0.0/16"
+	tablePrefixV6      = "2001:db8:1::/48"
+)
+
+// tableTime returns the time second seconds after the start of every routes
+// table test.
+func tableTime(second int) time.Time {
+	return time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC).
+		Add(time.Duration(second) * time.Second)
+}
+
+// announceTo has peer announce prefix with origin AS asn second seconds after
+// the start.
+func announceTo(
+	table *routetable.Table, prefix string, asn int, peer string, second int,
+) {
+	table.Announce(netip.MustParsePrefix(prefix), asn, routetable.Route{
+		Peer: peer, ASPath: []int{transitAS, asn}, Time: tableTime(second),
+	})
+}
+
+// newRoutesTestDatabase returns a new database in a temporary directory.
+func newRoutesTestDatabase(t *testing.T) *Database {
+	t.Helper()
+
+	db, err := New(&config.Config{StateDir: t.TempDir()}, logger.New())
+	if err != nil {
+		t.Fatalf("failed to create database: %v", err)
+	}
+
+	t.Cleanup(func() { _ = db.Close() })
+
+	return db
+}
+
+// writeTableChanges writes what table changed since the last call.
+func writeTableChanges(t *testing.T, db *Database, table *routetable.Table) {
+	t.Helper()
+
+	changes, newPeers := table.TakeChanges(math.MaxInt)
+
+	err := db.WriteRouteChanges(changes, newPeers)
+	if err != nil {
+		t.Fatalf("WriteRouteChanges: %v", err)
+	}
+}
+
+func TestReadRouteTableGivesBackWhatWasWritten(t *testing.T) {
+	t.Parallel()
+
+	db := newRoutesTestDatabase(t)
+	source := routetable.New(routetable.Params{})
+
+	// Two origins of one prefix: the routes of peerC and peerB are kept for
+	// asA, peerA keeps only its bit.
+	announceTo(source, tablePrefixV4, asA, peerA, 10)
+	announceTo(source, tablePrefixV4, asA, peerB, 20)
+	announceTo(source, tablePrefixV4, asA, peerC, 30)
+	announceTo(source, tablePrefixV4, asB, peerD, 40)
+	// Two routes with the same time.
+	announceTo(source, tablePrefixV6, asA, peerA, 50)
+	announceTo(source, tablePrefixV6, asA, peerC, 50)
+	// An origin whose kept routes were withdrawn while peerA carries it.
+	announceTo(source, tablePrefixShortV4, asB, peerA, 10)
+	announceTo(source, tablePrefixShortV4, asB, peerB, 20)
+	announceTo(source, tablePrefixShortV4, asB, peerC, 30)
+	source.Withdraw(netip.MustParsePrefix(tablePrefixShortV4), peerB)
+	source.Withdraw(netip.MustParsePrefix(tablePrefixShortV4), peerC)
+
+	writeTableChanges(t, db, source)
+	checkReadTable(t, db, source)
+
+	// A peer moves to another origin, one goes down and a new one comes.
+	announceTo(source, tablePrefixV4, asB, peerA, 60)
+	source.PeerDown(peerC)
+	announceTo(source, tablePrefixV6, asB, peerE, 70)
+
+	writeTableChanges(t, db, source)
+	checkReadTable(t, db, source)
+}
+
+func TestReadRouteTableGivesATableThatTakesLaterUpdates(t *testing.T) {
+	t.Parallel()
+
+	db := newRoutesTestDatabase(t)
+	source := routetable.New(routetable.Params{})
+
+	// peerA and peerB carry tablePrefixV4, peerB and peerC tablePrefixV6.
+	announceTo(source, tablePrefixV4, asA, peerA, 10)
+	announceTo(source, tablePrefixV4, asA, peerB, 20)
+	announceTo(source, tablePrefixV6, asA, peerB, 30)
+	announceTo(source, tablePrefixV6, asA, peerC, 40)
+
+	writeTableChanges(t, db, source)
+	read := checkReadTable(t, db, source)
+
+	// The same updates to both tables: a loaded peer withdraws a prefix,
+	// another one goes down, and a new peer comes.
+	for _, table := range []*routetable.Table{source, read} {
+		table.Withdraw(netip.MustParsePrefix(tablePrefixV4), peerA)
+		table.PeerDown(peerB)
+		announceTo(table, tablePrefixShortV4, asB, peerD, 50)
+	}
+
+	checkSameReads(t, read, source)
+}
+
+func TestReadRouteTableFailsOnAPeerWithoutANumber(t *testing.T) {
+	t.Parallel()
+
+	db := newRoutesTestDatabase(t)
+	table := routetable.New(routetable.Params{})
+
+	// Only the routes of peerB and peerC are kept, so only the peer set names
+	// peerA.
+	announceTo(table, tablePrefixV4, asA, peerA, 10)
+	announceTo(table, tablePrefixV4, asA, peerB, 20)
+	announceTo(table, tablePrefixV4, asA, peerC, 30)
+	writeTableChanges(t, db, table)
+
+	_, err := db.db.ExecContext(t.Context(),
+		"DELETE FROM route_peers WHERE address = ?", peerA)
+	if err != nil {
+		t.Fatalf("deleting a route peer: %v", err)
+	}
+
+	_, err = db.ReadRouteTable(t.Context())
+	if !errors.Is(err, errUnknownRoutePeer) {
+		t.Fatalf("ReadRouteTable = %v, want %v", err, errUnknownRoutePeer)
+	}
+}
+
+// checkReadTable checks that the table ReadRouteTable reads answers every read
+// as want does, and that reading it recorded no change. It returns that table.
+func checkReadTable(
+	t *testing.T, db *Database, want *routetable.Table,
+) *routetable.Table {
+	t.Helper()
+
+	got, err := db.ReadRouteTable(t.Context())
+	if err != nil {
+		t.Fatalf("ReadRouteTable: %v", err)
+	}
+
+	checkSameReads(t, got, want)
+
+	changes, newPeers := got.TakeChanges(math.MaxInt)
+	if len(changes) != 0 || len(newPeers) != 0 {
+		t.Errorf("the read recorded changes %+v and new peers %v, want none",
+			changes, newPeers)
+	}
+
+	return got
+}
+
+// checkSameReads checks that got answers every read as want does, for the
+// prefixes and the origin ASes the tests use.
+func checkSameReads(t *testing.T, got, want *routetable.Table) {
+	t.Helper()
+
+	if !reflect.DeepEqual(got.Counts(), want.Counts()) {
+		t.Errorf("Counts() = %+v, want %+v", got.Counts(), want.Counts())
+	}
+
+	for _, text := range []string{
+		tablePrefixV4, tablePrefixShortV4, tablePrefixV6,
+	} {
+		prefix := netip.MustParsePrefix(text)
+
+		if !reflect.DeepEqual(got.Origins(prefix), want.Origins(prefix)) {
+			t.Errorf("Origins(%s) = %+v, want %+v",
+				prefix, got.Origins(prefix), want.Origins(prefix))
+		}
+
+		gotSummary, gotFound := got.Lookup(prefix.Addr())
+		wantSummary, wantFound := want.Lookup(prefix.Addr())
+
+		if gotFound != wantFound || !reflect.DeepEqual(gotSummary, wantSummary) {
+			t.Errorf("Lookup(%s) = %+v, %v; want %+v, %v", prefix.Addr(),
+				gotSummary, gotFound, wantSummary, wantFound)
+		}
+
+		ipVersion := ipVersionV6
+		if prefix.Addr().Is4() {
+			ipVersion = ipVersionV4
+		}
+
+		gotSample := sortedSample(got, ipVersion, prefix.Bits())
+		wantSample := sortedSample(want, ipVersion, prefix.Bits())
+
+		if !reflect.DeepEqual(gotSample, wantSample) {
+			t.Errorf("SamplePrefixes(%d, %d) = %+v, want %+v",
+				ipVersion, prefix.Bits(), gotSample, wantSample)
+		}
+	}
+
+	for _, asn := range []int{asA, asB} {
+		if !reflect.DeepEqual(got.OriginatedPrefixes(asn),
+			want.OriginatedPrefixes(asn)) {
+			t.Errorf("OriginatedPrefixes(%d) = %+v, want %+v", asn,
+				got.OriginatedPrefixes(asn), want.OriginatedPrefixes(asn))
+		}
+	}
+}
+
+// sortedSample returns every prefix of the IP version and mask length in
+// table, in prefix order.
+func sortedSample(
+	table *routetable.Table, ipVersion, maskLength int,
+) []routetable.PrefixSummary {
+	// More than the tests have of any mask length.
+	const all = 100
+
+	sample := table.SamplePrefixes(ipVersion, maskLength, all)
+	slices.SortFunc(sample, func(a, b routetable.PrefixSummary) int {
+		return a.Prefix.Compare(b.Prefix)
+	})
+
+	return sample
+}
+
+func TestWriteRouteChangesDeletesTheRowOfARemovedPair(t *testing.T) {
+	t.Parallel()
+
+	db := newRoutesTestDatabase(t)
+	table := routetable.New(routetable.Params{})
+
+	announceTo(table, tablePrefixV4, asA, peerA, 10)
+	announceTo(table, tablePrefixV4, asB, peerB, 20)
+	writeTableChanges(t, db, table)
+
+	// The prefix's only peer with asA withdraws it.
+	table.Withdraw(netip.MustParsePrefix(tablePrefixV4), peerA)
+	writeTableChanges(t, db, table)
+
+	var origins []int
+
+	rows, err := db.db.QueryContext(t.Context(),
+		"SELECT origin_asn FROM routes WHERE prefix = ?", tablePrefixV4)
+	if err != nil {
+		t.Fatalf("reading the routes: %v", err)
+	}
+	defer func() { _ = rows.Close() }()
+
+	for rows.Next() {
+		var origin int
+
+		err := rows.Scan(&origin)
+		if err != nil {
+			t.Fatalf("reading a route: %v", err)
+		}
+
+		origins = append(origins, origin)
+	}
+
+	err = rows.Err()
+	if err != nil {
+		t.Fatalf("reading the routes: %v", err)
+	}
+
+	if !slices.Equal(origins, []int{asB}) {
+		t.Errorf("rows of %s have origins %v, want only %d",
+			tablePrefixV4, origins, asB)
+	}
+}
+
+// routePeer is one row of route_peers.
+type routePeer struct {
+	number  uint
+	address string
+}
+
+// readRoutePeerRows returns the rows of route_peers in number order.
+func readRoutePeerRows(t *testing.T, db *Database) []routePeer {
+	t.Helper()
+
+	rows, err := db.db.QueryContext(t.Context(),
+		"SELECT number, address FROM route_peers ORDER BY number")
+	if err != nil {
+		t.Fatalf("reading the route peers: %v", err)
+	}
+	defer func() { _ = rows.Close() }()
+
+	var peers []routePeer
+
+	for rows.Next() {
+		var peer routePeer
+
+		err := rows.Scan(&peer.number, &peer.address)
+		if err != nil {
+			t.Fatalf("reading a route peer: %v", err)
+		}
+
+		peers = append(peers, peer)
+	}
+
+	err = rows.Err()
+	if err != nil {
+		t.Fatalf("reading the route peers: %v", err)
+	}
+
+	return peers
+}
+
+func TestWriteRouteChangesStoresNewPeers(t *testing.T) {
+	t.Parallel()
+
+	db := newRoutesTestDatabase(t)
+	table := routetable.New(routetable.Params{})
+
+	announceTo(table, tablePrefixV4, asA, peerA, 10)
+	announceTo(table, tablePrefixV6, asA, peerC, 20)
+	writeTableChanges(t, db, table)
+
+	// peerA is stored already; peerB is new.
+	announceTo(table, tablePrefixV4, asA, peerA, 30)
+	announceTo(table, tablePrefixV4, asA, peerB, 40)
+	writeTableChanges(t, db, table)
+
+	want := []routePeer{{0, peerA}, {1, peerC}, {2, peerB}}
+
+	got := readRoutePeerRows(t, db)
+	if !slices.Equal(got, want) {
+		t.Errorf("route_peers = %+v, want %+v", got, want)
+	}
+}
+
+func TestWriteRouteChangesWritesNothingWhenItFails(t *testing.T) {
+	t.Parallel()
+
+	db := newRoutesTestDatabase(t)
+	table := routetable.New(routetable.Params{})
+
+	announceTo(table, tablePrefixV4, asA, peerA, 10)
+	announceTo(table, tablePrefixV6, asA, peerB, 20)
+
+	// peerB's number is missing, so its kept route cannot be written.
+	changes, newPeers := table.TakeChanges(math.MaxInt)
+	delete(newPeers, peerB)
+
+	err := db.WriteRouteChanges(changes, newPeers)
+	if !errors.Is(err, errUnknownRoutePeer) {
+		t.Fatalf("WriteRouteChanges = %v, want %v", err, errUnknownRoutePeer)
+	}
+
+	peers := readRoutePeerRows(t, db)
+	if len(peers) != 0 {
+		t.Errorf("route_peers = %+v after the failed write, want none", peers)
+	}
+
+	var routes int
+
+	err = db.db.QueryRowContext(t.Context(),
+		"SELECT COUNT(*) FROM routes").Scan(&routes)
+	if err != nil {
+		t.Fatalf("counting the routes: %v", err)
+	}
+
+	if routes != 0 {
+		t.Errorf("routes holds %d rows after the failed write, want none", routes)
+	}
+}
+
+// Parameters for BenchmarkReadRouteTable.
+const (
+	// benchTablePairs is how many prefix and origin AS pairs the table holds:
+	// one origin AS for each prefix. Every benchTableV6Every-th prefix is an
+	// IPv6 /48, the others IPv4 /24s.
+	benchTablePairs   = 200000
+	benchTableV6Every = 5
+	// benchTablePeers is how many peers there are, about as many as RIS has.
+	// benchTablePeersPerPair is how many of them, picked at random, carry each
+	// pair: RIS's 460 million routes over 1.4 million pairs.
+	benchTablePeers        = 1400
+	benchTablePeersPerPair = 330
+	// benchTableBatch is how many changes each write holds.
+	benchTableBatch = 10000
+)
+
+// BenchmarkReadRouteTable times the read of a routing table of 200,000 prefix
+// and origin AS pairs, each carried by 330 of 1,400 peers and keeping two
+// routes. It writes that table first. It reports the bytes of the database
+// file per pair. It is run by hand, not by make test, for example in the test
+// image:
+//
+//	go test -run '^$' -bench ReadRouteTable -benchtime 5x ./internal/database/
+func BenchmarkReadRouteTable(b *testing.B) {
+	db, err := New(&config.Config{StateDir: b.TempDir()}, logger.New())
+	if err != nil {
+		b.Fatalf("failed to create database: %v", err)
+	}
+	defer func() { _ = db.Close() }()
+
+	//nolint:gosec // The benchmark needs the same table every run.
+	rng := rand.New(rand.NewPCG(1, 2))
+
+	newPeers := make(map[string]uint, benchTablePeers)
+	for number := range uint(benchTablePeers) {
+		newPeers[benchPeerAddress(number)] = number
+	}
+
+	for first := 0; first < benchTablePairs; first += benchTableBatch {
+		changes := make([]routetable.Change, 0, benchTableBatch)
+		for pair := first; pair < first+benchTableBatch; pair++ {
+			changes = append(changes, benchTableChange(rng, pair))
+		}
+
+		err := db.WriteRouteChanges(changes, newPeers)
+		if err != nil {
+			b.Fatalf("WriteRouteChanges: %v", err)
+		}
+
+		newPeers = nil
+	}
+
+	// The WAL is copied into the database file, whose size is reported.
+	err = db.Checkpoint(b.Context())
+	if err != nil {
+		b.Fatalf("Checkpoint: %v", err)
+	}
+
+	file, err := os.Stat(db.path)
+	if err != nil {
+		b.Fatalf("reading the database file's size: %v", err)
+	}
+
+	for b.Loop() {
+		table, err := db.ReadRouteTable(b.Context())
+		if err != nil {
+			b.Fatalf("ReadRouteTable: %v", err)
+		}
+
+		counts := table.Counts()
+		if counts.IPv4Prefixes+counts.IPv6Prefixes != benchTablePairs {
+			b.Fatalf("read %d prefixes, want %d",
+				counts.IPv4Prefixes+counts.IPv6Prefixes, benchTablePairs)
+		}
+	}
+
+	// After the loop: b.Loop drops the metrics reported before it starts.
+	b.ReportMetric(float64(file.Size())/benchTablePairs, "file-bytes/pair")
+}
+
+// benchPeerAddress returns the address of the peer numbered number.
+func benchPeerAddress(number uint) string {
+	const addressesPerBlock = 256
+
+	return fmt.Sprintf("198.18.%d.%d",
+		number/addressesPerBlock, number%addressesPerBlock)
+}
+
+// benchTableChange returns the content of the pair numbered pair: its prefix,
+// its origin AS, benchTablePeersPerPair peers picked at random, and the routes
+// of two of them, each with an AS path of four ASes.
+func benchTableChange(rng *rand.Rand, pair int) routetable.Change {
+	const (
+		firstOrigin  = 64500
+		transitASes  = 50000
+		olderSeconds = 10
+	)
+
+	text := fmt.Sprintf("%d.%d.%d.0/24", 1+pair>>16, (pair>>8)%256, pair%256)
+	if pair%benchTableV6Every == 0 {
+		text = fmt.Sprintf("2001:%x:%x::/48", pair>>16, pair%65536)
+	}
+
+	prefix := netip.MustParsePrefix(text)
+	origin := firstOrigin + pair%transitASes
+	peers := bitset.New(benchTablePeers)
+
+	for peers.Count() < benchTablePeersPerPair {
+		peers.Set(rng.UintN(benchTablePeers))
+	}
+
+	change := routetable.Change{Prefix: prefix, OriginASN: origin, Peers: peers}
+
+	// The routes of the first two peers, newest first.
+	for i, number := range slices.Collect(peers.EachSet())[:keptRoutesPerRow] {
+		change.Routes = append(change.Routes, routetable.Route{
+			Peer: benchPeerAddress(number),
+			ASPath: []int{
+				firstOrigin + rng.IntN(transitASes),
+				firstOrigin + rng.IntN(transitASes),
+				firstOrigin + rng.IntN(transitASes),
+				origin,
+			},
+			Time: tableTime(pair - i*olderSeconds),
+		})
+	}
+
+	return change
+}

internal/database/schema.sql

@@ -27,21 +27,18 @@ CREATE TABLE IF NOT EXISTS asns (
 	whois_updated_at DATETIME    -- when we last fetched WHOIS data
 );
 
--- IPv4 prefixes table
+-- IPv4 prefixes table. Every read and write finds a row by its prefix, so the
+-- rows are stored by prefix (WITHOUT ROWID) and need no index of their own.
 CREATE TABLE IF NOT EXISTS prefixes_v4 (
-	id TEXT PRIMARY KEY,
+	prefix TEXT PRIMARY KEY,
-	prefix TEXT UNIQUE NOT NULL,
+	first_seen DATETIME NOT NULL
-	first_seen DATETIME NOT NULL,
+) WITHOUT ROWID;
-	last_seen DATETIME NOT NULL
-);
 
--- IPv6 prefixes table
+-- IPv6 prefixes table, stored as prefixes_v4 is.
 CREATE TABLE IF NOT EXISTS prefixes_v6 (
-	id TEXT PRIMARY KEY,
+	prefix TEXT PRIMARY KEY,
-	prefix TEXT UNIQUE NOT NULL,
+	first_seen DATETIME NOT NULL
-	first_seen DATETIME NOT NULL,
+) WITHOUT ROWID;
-	last_seen DATETIME NOT NULL
-);
 
 CREATE TABLE IF NOT EXISTS announcements (
 	id TEXT PRIMARY KEY,
@@ -75,11 +72,6 @@ CREATE TABLE IF NOT EXISTS bgp_peers (
 	last_message_type TEXT
 );
 
--- Indexes for prefixes_v4 table
-CREATE INDEX IF NOT EXISTS idx_prefixes_v4_prefix ON prefixes_v4(prefix);
-
--- Indexes for prefixes_v6 table
-CREATE INDEX IF NOT EXISTS idx_prefixes_v6_prefix ON prefixes_v6(prefix);
 CREATE INDEX IF NOT EXISTS idx_announcements_timestamp ON announcements(timestamp);
 CREATE INDEX IF NOT EXISTS idx_announcements_prefix_id ON announcements(prefix_id);
 CREATE INDEX IF NOT EXISTS idx_announcements_peer_asn ON announcements(peer_asn);
@@ -97,9 +89,14 @@ CREATE INDEX IF NOT EXISTS idx_bgp_peers_asn ON bgp_peers(peer_asn);
 CREATE INDEX IF NOT EXISTS idx_bgp_peers_last_seen ON bgp_peers(last_seen);
 CREATE INDEX IF NOT EXISTS idx_bgp_peers_ip ON bgp_peers(peer_ip);
 
--- IPv4 routing table maintained by PrefixHandler
+-- IPv4 routing table maintained by PrefixHandler. A route write changes the
+-- table and its entry in every index on a column the write sets, so the table
+-- has only the indexes that the queries named with them use.
 CREATE TABLE IF NOT EXISTS live_routes_v4 (
-	id TEXT PRIMARY KEY,
+	-- The row id, which the prefix-length pages' sample draws through the
+	-- mask_length index. Declared as a column, it stays the same for as long
+	-- as the route is live, also through a VACUUM.
+	id INTEGER PRIMARY KEY,
 	prefix TEXT NOT NULL,
 	mask_length INTEGER NOT NULL, -- CIDR mask length (0-32)
 	origin_asn INTEGER NOT NULL,
@@ -107,12 +104,15 @@ CREATE TABLE IF NOT EXISTS live_routes_v4 (
 	as_path TEXT NOT NULL, -- JSON array
 	next_hop TEXT NOT NULL,
 	last_updated DATETIME NOT NULL,
-	UNIQUE(prefix, origin_asn, peer_ip)
+	-- A peer has at most one route for a prefix. This key finds the route a
+	-- write updates or deletes and, peer_ip coming first, all of a peer's
+	-- routes when its BGP session goes down.
+	UNIQUE(peer_ip, prefix)
 );
 
--- IPv6 routing table maintained by PrefixHandler
+-- IPv6 routing table maintained by PrefixHandler, as live_routes_v4.
 CREATE TABLE IF NOT EXISTS live_routes_v6 (
-	id TEXT PRIMARY KEY,
+	id INTEGER PRIMARY KEY,
 	prefix TEXT NOT NULL,
 	mask_length INTEGER NOT NULL, -- CIDR mask length (0-128)
 	origin_asn INTEGER NOT NULL,
@@ -120,21 +120,59 @@ CREATE TABLE IF NOT EXISTS live_routes_v6 (
 	as_path TEXT NOT NULL, -- JSON array
 	next_hop TEXT NOT NULL,
 	last_updated DATETIME NOT NULL,
-	UNIQUE(prefix, origin_asn, peer_ip)
+	UNIQUE(peer_ip, prefix)
 );
 
--- Indexes for live_routes_v4 table
+-- Indexes for both live route tables, and the queries that use each.
-CREATE INDEX IF NOT EXISTS idx_live_routes_v4_prefix ON live_routes_v4(prefix);
+--
+-- prefix, last_updated: a prefix's routes, for the prefix page and its API
+-- answer, the IP lookup and its peer count, and the check whether a prefix
+-- still has a live route; and a prefix's newest route for the prefix-length
+-- pages and the AS page, in one search whatever the number of peers.
+CREATE INDEX IF NOT EXISTS idx_live_routes_v4_prefix_last_updated ON live_routes_v4(prefix, last_updated);
+CREATE INDEX IF NOT EXISTS idx_live_routes_v6_prefix_last_updated ON live_routes_v6(prefix, last_updated);
+-- mask_length: the prefix-length pages' random sample, the row id range of a
+-- mask length and the first route of it at or after a row id.
 CREATE INDEX IF NOT EXISTS idx_live_routes_v4_mask_length ON live_routes_v4(mask_length);
-CREATE INDEX IF NOT EXISTS idx_live_routes_v4_origin_asn ON live_routes_v4(origin_asn);
-CREATE INDEX IF NOT EXISTS idx_live_routes_v4_last_updated ON live_routes_v4(last_updated);
--- Index to optimize prefix distribution queries
-CREATE INDEX IF NOT EXISTS idx_live_routes_v4_mask_prefix ON live_routes_v4(mask_length, prefix);
-
--- Indexes for live_routes_v6 table
-CREATE INDEX IF NOT EXISTS idx_live_routes_v6_prefix ON live_routes_v6(prefix);
 CREATE INDEX IF NOT EXISTS idx_live_routes_v6_mask_length ON live_routes_v6(mask_length);
-CREATE INDEX IF NOT EXISTS idx_live_routes_v6_origin_asn ON live_routes_v6(origin_asn);
+-- mask_length, prefix: the prefix-length pages' list of a mask length's
-CREATE INDEX IF NOT EXISTS idx_live_routes_v6_last_updated ON live_routes_v6(last_updated);
+-- prefixes in text order, and the prefix distribution read at startup.
--- Index to optimize prefix distribution queries
+CREATE INDEX IF NOT EXISTS idx_live_routes_v4_mask_prefix ON live_routes_v4(mask_length, prefix);
 CREATE INDEX IF NOT EXISTS idx_live_routes_v6_mask_prefix ON live_routes_v6(mask_length, prefix);
+-- origin_asn, prefix: the prefixes an AS originates, each in one search
+-- whatever the number of peers, for the AS page and its API answer.
+CREATE INDEX IF NOT EXISTS idx_live_routes_v4_origin_prefix ON live_routes_v4(origin_asn, prefix);
+CREATE INDEX IF NOT EXISTS idx_live_routes_v6_origin_prefix ON live_routes_v6(origin_asn, prefix);
+-- last_updated: the oldest and newest route times in /api/v1/stats and
+-- /status.json.
+CREATE INDEX IF NOT EXISTS idx_live_routes_v4_last_updated ON live_routes_v4(last_updated);
+CREATE INDEX IF NOT EXISTS idx_live_routes_v6_last_updated ON live_routes_v6(last_updated);
+
+-- The whole routing table, as internal/routetable keeps it in memory: one row
+-- for each prefix and origin AS that some peer carries. Startup reads every
+-- row, and the rows of the pairs that change are written back.
+CREATE TABLE IF NOT EXISTS routes (
+	prefix TEXT NOT NULL, -- in the text form Go's net/netip prints
+	origin_asn INTEGER NOT NULL,
+	-- The peers whose current route to the prefix has this origin AS: a set
+	-- of their numbers in route_peers, as the bitset package's MarshalBinary
+	-- writes it.
+	peers BLOB NOT NULL,
+	-- The newest of those routes and the one before it, the two kept in full.
+	-- Each has its peer's number in route_peers, its AS path as a JSON array,
+	-- and its time in seconds since 1970, UTC. A route that is not kept has
+	-- NULL in its three columns.
+	newest_peer INTEGER,
+	newest_as_path TEXT,
+	newest_time INTEGER,
+	older_peer INTEGER,
+	older_as_path TEXT,
+	older_time INTEGER,
+	PRIMARY KEY (prefix, origin_asn)
+) WITHOUT ROWID;
+
+-- The peers of the routes table, each with the number it has there.
+CREATE TABLE IF NOT EXISTS route_peers (
+	number INTEGER PRIMARY KEY,
+	address TEXT NOT NULL UNIQUE
+);

internal/database/slowquery.go

@@ -18,30 +18,12 @@ func logSlowQuery(logger *logger.Logger, query string, start time.Time) {
 	}
 }
 
-// queryRow wraps QueryRow with slow query logging
+// exec runs a statement on the writing connection, with slow query logging
-// nolint:unused // kept for consistency with other query wrappers
+func (d *Database) exec(query string, args ...any) error {
-func (d *Database) queryRow(query string, args ...interface{}) *sql.Row {
 	start := time.Now()
 	defer logSlowQuery(d.logger, query, start)
 
-	return d.db.QueryRow(query, args...)
+	_, err := d.writer.ExecContext(context.Background(), query, args...)
-}
-
-// query wraps Query with slow query logging
-// nolint:unused // kept for future use to ensure all queries go through slow query logging
-func (d *Database) query(query string, args ...interface{}) (*sql.Rows, error) {
-	start := time.Now()
-	defer logSlowQuery(d.logger, query, start)
-
-	return d.db.Query(query, args...)
-}
-
-// exec wraps Exec with slow query logging
-func (d *Database) exec(query string, args ...interface{}) error {
-	start := time.Now()
-	defer logSlowQuery(d.logger, query, start)
-
-	_, err := d.db.Exec(query, args...)
 
 	return err
 }
@@ -49,51 +31,51 @@ func (d *Database) exec(query string, args ...interface{}) error {
 // loggingTx wraps sql.Tx to log slow queries
 type loggingTx struct {
 	*sql.Tx
+
 	logger *logger.Logger
 }
 
 // QueryRow wraps sql.Tx.QueryRow to log slow queries
-func (tx *loggingTx) QueryRow(query string, args ...interface{}) *sql.Row {
+func (tx *loggingTx) QueryRow(query string, args ...any) *sql.Row {
 	start := time.Now()
 	defer logSlowQuery(tx.logger, query, start)
 
-	return tx.Tx.QueryRow(query, args...)
+	return tx.Tx.QueryRowContext(context.Background(), query, args...)
-}
-
-// Query wraps sql.Tx.Query to log slow queries
-func (tx *loggingTx) Query(query string, args ...interface{}) (*sql.Rows, error) {
-	start := time.Now()
-	defer logSlowQuery(tx.logger, query, start)
-
-	return tx.Tx.Query(query, args...)
-}
-
-// QueryContext wraps sql.Tx.QueryContext to log slow queries
-func (tx *loggingTx) QueryContext(ctx context.Context, query string, args ...interface{}) (*sql.Rows, error) {
-	start := time.Now()
-	defer logSlowQuery(tx.logger, query, start)
-
-	return tx.Tx.QueryContext(ctx, query, args...)
 }
 
 // QueryRowContext wraps sql.Tx.QueryRowContext to log slow queries
-func (tx *loggingTx) QueryRowContext(ctx context.Context, query string, args ...interface{}) *sql.Row {
+func (tx *loggingTx) QueryRowContext(
+	ctx context.Context, query string, args ...any,
+) *sql.Row {
 	start := time.Now()
 	defer logSlowQuery(tx.logger, query, start)
 
 	return tx.Tx.QueryRowContext(ctx, query, args...)
 }
 
-// Exec wraps sql.Tx.Exec to log slow queries
+// QueryContext wraps sql.Tx.QueryContext to log slow queries
-func (tx *loggingTx) Exec(query string, args ...interface{}) (sql.Result, error) {
+func (tx *loggingTx) QueryContext(
+	ctx context.Context, query string, args ...any,
+) (*sql.Rows, error) {
 	start := time.Now()
 	defer logSlowQuery(tx.logger, query, start)
 
-	return tx.Tx.Exec(query, args...)
+	//nolint:sqlclosecheck // returns the rows, which the caller closes
+	return tx.Tx.QueryContext(ctx, query, args...)
+}
+
+// Exec wraps sql.Tx.Exec to log slow queries
+func (tx *loggingTx) Exec(query string, args ...any) (sql.Result, error) {
+	start := time.Now()
+	defer logSlowQuery(tx.logger, query, start)
+
+	return tx.Tx.ExecContext(context.Background(), query, args...)
 }
 
 // ExecContext wraps sql.Tx.ExecContext to log slow queries
-func (tx *loggingTx) ExecContext(ctx context.Context, query string, args ...interface{}) (sql.Result, error) {
+func (tx *loggingTx) ExecContext(
+	ctx context.Context, query string, args ...any,
+) (sql.Result, error) {
 	start := time.Now()
 	defer logSlowQuery(tx.logger, query, start)
 

internal/database/utils.go

@@ -27,6 +27,7 @@ func prefixMaskLength(prefix string) (int, error) {
 	if err != nil {
 		return 0, fmt.Errorf("invalid prefix %s: %w", prefix, err)
 	}
+
 	maskLength, _ := network.Mask.Size()
 
 	return maskLength, nil

internal/logger/logger.go

@@ -74,6 +74,7 @@ func getSourceAttrs() []slog.Attr {
 
 	// Get the function name
 	fn := runtime.FuncForPC(pc)
+
 	var funcName string
 	if fn != nil {
 		funcName = filepath.Base(fn.Name())

internal/metrics/metrics.go

@@ -34,7 +34,7 @@ type Tracker struct {
 	// Announcement/withdrawal metrics
 	announcementCounter metrics.Counter
 	withdrawalCounter   metrics.Counter
-	churnRate           metrics.Meter // combined announcements + withdrawals per second
+	churnPerMinute      minuteCounter // prefix announcements and withdrawals
 
 	// BGP peer tracking
 	bgpPeerCount atomic.Int32
@@ -56,23 +56,29 @@ func New() *Tracker {
 		ipv6UpdateRate:      metrics.NewMeter(),
 		announcementCounter: metrics.NewCounter(),
 		withdrawalCounter:   metrics.NewCounter(),
-		churnRate:           metrics.NewMeter(),
+		churnPerMinute:      minuteCounter{now: time.Now},
 	}
 }
 
-// SetConnected updates the connection status
+// SetConnected updates the connection status. Every connection after the
+// first counts one reconnect.
 func (t *Tracker) SetConnected(connected bool) {
-	wasConnected := t.isConnected.Swap(connected)
+	t.isConnected.Store(connected)
-	if connected {
+
+	if !connected {
+		return
+	}
+
 	t.mu.Lock()
+	// connectedSince is zero until the first connection.
+	connectedBefore := !t.connectedSince.IsZero()
 	t.connectedSince = time.Now()
 	t.mu.Unlock()
-		// Increment reconnect count (but not for the initial connection)
+
-		if wasConnected || t.reconnectCount.Load() > 0 {
+	if connectedBefore {
 		t.reconnectCount.Add(1)
 	}
 }
-}
 
 // GetReconnectCount returns the number of reconnections since startup
 func (t *Tracker) GetReconnectCount() uint64 {
@@ -115,9 +121,11 @@ func (t *Tracker) GetStreamMetrics() StreamMetrics {
 	if msgCount >= 0 {
 		totalMessages = uint64(msgCount)
 	}
+
 	if byteCount >= 0 {
 		totalBytes = uint64(byteCount)
 	}
+
 	if wireByteCount >= 0 {
 		totalWireBytes = uint64(wireByteCount)
 	}
@@ -148,20 +156,20 @@ func (t *Tracker) RecordIPv6Update() {
 // RecordAnnouncement records a route announcement
 func (t *Tracker) RecordAnnouncement() {
 	t.announcementCounter.Inc(1)
-	t.churnRate.Mark(1)
+	t.churnPerMinute.add()
 }
 
 // RecordWithdrawal records a route withdrawal
 func (t *Tracker) RecordWithdrawal() {
 	t.withdrawalCounter.Inc(1)
-	t.churnRate.Mark(1)
+	t.churnPerMinute.add()
 }
 
 // SetBGPPeerCount updates the current BGP peer count
 func (t *Tracker) SetBGPPeerCount(count int) {
 	// BGP peer count is always small (< 1000), so int32 is safe
-	if count > 0 && count < 1<<31 {
+	if count >= 0 && count < 1<<31 {
-		t.bgpPeerCount.Store(int32(count)) //nolint:gosec // count is validated
+		t.bgpPeerCount.Store(int32(count))
 	}
 }
 
@@ -190,9 +198,10 @@ func (t *Tracker) GetWithdrawalCount() uint64 {
 	return uint64(count)
 }
 
-// GetChurnRate returns the route churn rate per second
+// GetChurnPerMinute returns the number of prefix announcements and
-func (t *Tracker) GetChurnRate() float64 {
+// withdrawals, from all peers, in the last full clock minute.
-	return t.churnRate.Rate1()
+func (t *Tracker) GetChurnPerMinute() int {
+	return t.churnPerMinute.lastFullMinute()
 }
 
 // GetRouteMetrics returns current route update metrics
@@ -209,7 +218,8 @@ type StreamMetrics struct {
 	TotalMessages uint64
 	// TotalBytes is the total number of decompressed bytes received since startup
 	TotalBytes uint64
-	// TotalWireBytes is the total number of bytes received on the wire (before decompression)
+	// TotalWireBytes is the total number of bytes received on the wire (before
+	// decompression)
 	TotalWireBytes uint64
 	// ConnectedSince is the time when the current connection was established
 	ConnectedSince time.Time
@@ -217,9 +227,11 @@ type StreamMetrics struct {
 	Connected bool
 	// MessagesPerSec is the rate of messages received per second (1-minute average)
 	MessagesPerSec float64
-	// BitsPerSec is the rate of decompressed bits received per second (1-minute average)
+	// BitsPerSec is the rate of decompressed bits received per second (1-minute
+	// average)
 	BitsPerSec float64
-	// WireBitsPerSec is the rate of bits received on the wire per second (1-minute average)
+	// WireBitsPerSec is the rate of bits received on the wire per second
+	// (1-minute average)
 	WireBitsPerSec float64
 	// ReconnectCount is the number of reconnections since startup
 	ReconnectCount uint64

internal/metrics/metrics_test.go

@@ -0,0 +1,42 @@
+package metrics_test
+
+import (
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/metrics"
+)
+
+// TestReconnectCount connects three times with a disconnect between each, the
+// way the streamer does, and expects every connection after the first to count
+// one reconnect.
+func TestReconnectCount(t *testing.T) {
+	t.Parallel()
+
+	tracker := metrics.New()
+
+	tracker.SetConnected(true)
+	tracker.SetConnected(false)
+	tracker.SetConnected(true)
+	tracker.SetConnected(false)
+	tracker.SetConnected(true)
+
+	const wantReconnects = 2
+	if got := tracker.GetReconnectCount(); got != wantReconnects {
+		t.Fatalf("reconnect count = %d, want %d", got, wantReconnects)
+	}
+}
+
+// TestBGPPeerCountDropsToZero sets a peer count of 1 and then 0, the way the
+// streamer does when the last peer's session ends, and expects 0 back.
+func TestBGPPeerCountDropsToZero(t *testing.T) {
+	t.Parallel()
+
+	tracker := metrics.New()
+
+	tracker.SetBGPPeerCount(1)
+	tracker.SetBGPPeerCount(0)
+
+	if got := tracker.GetBGPPeerCount(); got != 0 {
+		t.Fatalf("peer count = %d after setting 0, want 0", got)
+	}
+}

internal/metrics/minutecounter.go

@@ -0,0 +1,59 @@
+package metrics
+
+import (
+	"sync"
+	"time"
+)
+
+// minuteCounter counts events in each clock minute and reports the count of
+// the last full minute. It is safe for concurrent use.
+type minuteCounter struct {
+	mu  sync.Mutex
+	now func() time.Time
+
+	// minute is the start of the minute that count belongs to.
+	minute time.Time
+	count  int
+	// lastMinuteCount is the count of the minute just before minute.
+	lastMinuteCount int
+}
+
+// add counts one event in the current minute.
+func (c *minuteCounter) add() {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	c.moveToCurrentMinute()
+	c.count++
+}
+
+// lastFullMinute returns the number of events in the minute that ended most
+// recently.
+func (c *minuteCounter) lastFullMinute() int {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+
+	c.moveToCurrentMinute()
+
+	return c.lastMinuteCount
+}
+
+// moveToCurrentMinute starts a new count when the clock has moved into
+// another minute. The count it ends becomes the last full minute's count when
+// its minute is the one just before the current minute. Otherwise the last
+// full minute had no events. The caller holds mu.
+func (c *minuteCounter) moveToCurrentMinute() {
+	minute := c.now().Truncate(time.Minute)
+	if minute.Equal(c.minute) {
+		return
+	}
+
+	if minute.Sub(c.minute) == time.Minute {
+		c.lastMinuteCount = c.count
+	} else {
+		c.lastMinuteCount = 0
+	}
+
+	c.minute = minute
+	c.count = 0
+}

internal/metrics/minutecounter_internal_test.go

@@ -0,0 +1,86 @@
+package metrics
+
+import (
+	"sync"
+	"testing"
+	"time"
+)
+
+// TestChurnPerMinuteCountsLastFullMinute announces and withdraws prefixes on
+// an injected clock and expects each minute's whole count once that minute
+// has ended, and 0 for a minute with none.
+func TestChurnPerMinuteCountsLastFullMinute(t *testing.T) {
+	t.Parallel()
+
+	now := time.Date(2026, 10, 26, 12, 0, 30, 0, time.UTC)
+	tracker := New()
+	tracker.churnPerMinute.now = func() time.Time { return now }
+
+	expectChurn := func(want int, why string) {
+		t.Helper()
+
+		got := tracker.GetChurnPerMinute()
+		if got != want {
+			t.Fatalf("at %s churn = %d, want %d: %s",
+				now.Format(time.TimeOnly), got, want, why)
+		}
+	}
+
+	const firstMinuteChurn = 2
+
+	tracker.RecordAnnouncement()
+	tracker.RecordWithdrawal()
+	expectChurn(0, "12:00 has not ended")
+
+	now = time.Date(2026, 10, 26, 12, 1, 59, 0, time.UTC)
+
+	tracker.RecordAnnouncement()
+	expectChurn(firstMinuteChurn, "12:00 had one announcement and one withdrawal")
+
+	now = time.Date(2026, 10, 26, 12, 2, 0, 0, time.UTC)
+
+	expectChurn(1, "12:01 had one announcement")
+
+	now = time.Date(2026, 10, 26, 12, 4, 0, 0, time.UTC)
+
+	expectChurn(0, "12:03 had none")
+
+	tracker.RecordWithdrawal()
+
+	now = time.Date(2026, 10, 26, 12, 6, 0, 0, time.UTC)
+
+	expectChurn(0, "12:05 had none, though 12:04 had one withdrawal")
+}
+
+// TestChurnPerMinuteCountsConcurrentRecords announces prefixes from several
+// goroutines at once and expects every one counted.
+func TestChurnPerMinuteCountsConcurrentRecords(t *testing.T) {
+	t.Parallel()
+
+	const (
+		goroutines          = 8
+		recordsPerGoroutine = 1000
+	)
+
+	now := time.Date(2026, 10, 26, 12, 0, 0, 0, time.UTC)
+	tracker := New()
+	tracker.churnPerMinute.now = func() time.Time { return now }
+
+	var wg sync.WaitGroup
+	for range goroutines {
+		wg.Go(func() {
+			for range recordsPerGoroutine {
+				tracker.RecordAnnouncement()
+			}
+		})
+	}
+
+	wg.Wait()
+
+	now = now.Add(time.Minute)
+
+	want := goroutines * recordsPerGoroutine
+	if got := tracker.GetChurnPerMinute(); got != want {
+		t.Fatalf("churn = %d, want %d", got, want)
+	}
+}

internal/ristypes/ris.go

@@ -1,55 +1,122 @@
-// Package ristypes defines the data structures for RIS Live BGP messages and announcements.
+// Package ristypes defines the data structures for RIS Live BGP messages and
+// announcements.
 package ristypes
 
 import (
 	"encoding/json"
+	"slices"
 	"time"
 )
 
-// ASPath represents a BGP AS path as a slice of AS numbers.
+// Message types and the session state the handlers act on, as the feed writes
-// It handles JSON unmarshaling of both simple arrays and nested AS sets,
+// them in a message's type and state fields.
-// flattening any nested structures into a single sequence of AS numbers.
+const (
-type ASPath []int
+	// MessageTypeUpdate is a BGP UPDATE: announcements and withdrawals.
+	MessageTypeUpdate = "UPDATE"
 
-// UnmarshalJSON implements the json.Unmarshaler interface for ASPath.
+	// MessageTypeState is a change in the state of the BGP session between
-// It handles both simple integer arrays [1, 2, 3] and nested AS sets
+	// the collector and the peer.
-// like [1, [2, 3], 4], flattening them into a single slice of integers.
+	MessageTypeState = "STATE"
-func (p *ASPath) UnmarshalJSON(data []byte) error {
-	// First try to unmarshal as a simple array of integers
-	var simple []int
-	if err := json.Unmarshal(data, &simple); err == nil {
-		*p = ASPath(simple)
 
-		return nil
+	// StateDown is the state of a STATE message whose session has ended: the
+	// collector no longer has any of the peer's routes, and sends no
+	// withdrawals for them.
+	StateDown = "down"
+)
+
+// ASPath is a BGP AS path as the feed writes it: an array of AS numbers in
+// which an AS set is an array of its members, such as [1, 2, [3, 4]]. A path
+// that holds AS 0 decodes as an empty one (UnmarshalJSON).
+type ASPath struct {
+	// Sequence is the path's AS numbers in order. A set of one member that
+	// ends the path counts as that AS and is its last element. The members of
+	// a set anywhere else are flattened into it.
+	Sequence []int
+
+	// Set holds the members of a set of several members that ends the path.
+	// Any of them may be the origin, and they are not neighbours of each
+	// other or of the AS before them.
+	Set []int
 }
 
-	// If that fails, unmarshal as array of interfaces and flatten
+// NoOrigin is the origin AS of a route whose origin is not known. AS 0 is
-	var raw []interface{}
+// reserved, so no AS has this number.
-	if err := json.Unmarshal(data, &raw); err != nil {
+const NoOrigin = 0
+
+// Origin returns the AS that originated the route: the last AS of the
+// sequence, or NoOrigin when the path ends in a set of several members or is
+// empty.
+func (p *ASPath) Origin() int {
+	if len(p.Set) > 0 || len(p.Sequence) == 0 {
+		return NoOrigin
+	}
+
+	return p.Sequence[len(p.Sequence)-1]
+}
+
+// UnmarshalJSON implements the json.Unmarshaler interface for ASPath. A path
+// that holds AS 0 anywhere, a member of an AS set included, decodes as an empty
+// path: RFC 7607 makes such a path malformed and its announcements
+// withdrawals. On an announcement with an empty path the prefix handler
+// deletes the peer's route, and the AS and peering handlers record nothing.
+func (p *ASPath) UnmarshalJSON(data []byte) error {
+	path, err := decodeASPath(data)
+	if err != nil {
 		return err
 	}
 
-	// Flatten the array
+	if slices.Contains(path.Sequence, 0) || slices.Contains(path.Set, 0) {
-	result := make([]int, 0)
+		path = ASPath{}
-	for _, item := range raw {
-		switch v := item.(type) {
-		case float64:
-			result = append(result, int(v))
-		case []interface{}:
-			// Nested array - flatten it
-			for _, nested := range v {
-				if num, ok := nested.(float64); ok {
-					result = append(result, int(num))
-				}
-			}
-		}
 	}
 
-	*p = ASPath(result)
+	*p = path
 
 	return nil
 }
 
+// decodeASPath decodes an AS path as the feed writes it.
+func decodeASPath(data []byte) (ASPath, error) {
+	// Most paths have no AS set and decode as a plain array of integers.
+	var sequence []int
+
+	err := json.Unmarshal(data, &sequence)
+	if err == nil {
+		return ASPath{Sequence: sequence}, nil
+	}
+
+	var raw []any
+
+	err = json.Unmarshal(data, &raw)
+	if err != nil {
+		return ASPath{}, err
+	}
+
+	path := ASPath{Sequence: make([]int, 0, len(raw))}
+
+	for i, item := range raw {
+		switch v := item.(type) {
+		case float64:
+			path.Sequence = append(path.Sequence, int(v))
+		case []any:
+			set := make([]int, 0, len(v))
+
+			for _, member := range v {
+				if num, ok := member.(float64); ok {
+					set = append(set, int(num))
+				}
+			}
+
+			if i == len(raw)-1 && len(set) > 1 {
+				path.Set = set
+			} else {
+				path.Sequence = append(path.Sequence, set...)
+			}
+		}
+	}
+
+	return path, nil
+}
+
 // RISLiveMessage represents the outer wrapper message from the RIPE RIS Live stream.
 // Each message contains a Type field indicating the message type and a Data field
 // containing the actual BGP message payload.
@@ -62,6 +129,8 @@ type RISLiveMessage struct {
 // It contains metadata about the BGP session (peer, ASN, host) along with
 // the actual BGP update data including AS path, communities, announcements,
 // and withdrawals.
+//
+//nolint:tagliatelle // field names are the RIS Live feed's
 type RISMessage struct {
 	Type            string    `json:"type"`
 	Timestamp       float64   `json:"timestamp"`
@@ -73,7 +142,7 @@ type RISMessage struct {
 	RRC             string    `json:"rrc,omitempty"`
 	MrtTime         float64   `json:"mrt_time,omitempty"`
 	SocketTime      float64   `json:"socket_time,omitempty"`
-	Path            ASPath    `json:"path,omitempty"`
+	Path            ASPath    `json:"path,omitzero"`
 	// Community and Raw are present in the feed but read by no handler.
 	// They are the largest fields on a message that lives in up to four
 	// handler queues, so json:"-" keeps them out of the decoded message
@@ -85,11 +154,15 @@ type RISMessage struct {
 	Announcements []RISAnnouncement `json:"announcements,omitempty"`
 	Withdrawals   []string          `json:"withdrawals,omitempty"`
 	Raw           string            `json:"-"`
+	// State is set on STATE messages: the session's new state, such as StateDown.
+	State string `json:"state,omitempty"`
 }
 
 // RISAnnouncement represents a BGP route announcement within a RIS message.
 // It contains the next hop IP address and the list of prefixes being announced
 // via that next hop.
+//
+//nolint:tagliatelle // field names are the RIS Live feed's
 type RISAnnouncement struct {
 	NextHop  string   `json:"next_hop"`
 	Prefixes []string `json:"prefixes"`

internal/ristypes/ris_test.go

@@ -1,10 +1,13 @@
-package ristypes
+package ristypes_test
 
 import (
 	"bufio"
 	"encoding/json"
 	"os"
+	"slices"
 	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/ristypes"
 )
 
 // messageExamplesPath is the captured RIS Live feed used as decode fixtures,
@@ -15,44 +18,58 @@ const messageExamplesPath = "../../docs/message-examples.json"
 // streamer does and checks that the fields no handler reads (Community, Raw)
 // stay empty while the fields handlers use (Path, Announcements) still decode.
 func TestDecodeDropsCommunityAndRaw(t *testing.T) {
+	t.Parallel()
+
 	f, err := os.Open(messageExamplesPath)
 	if err != nil {
 		t.Fatalf("open fixtures: %v", err)
 	}
-	defer f.Close()
+
+	defer func() {
+		_ = f.Close()
+	}()
 
 	var messages, withPath, withAnnouncements int
 
 	scanner := bufio.NewScanner(f)
 	scanner.Buffer(make([]byte, 0, 64*1024), 1024*1024)
+
 	for scanner.Scan() {
 		line := scanner.Bytes()
 		if len(line) == 0 {
 			continue
 		}
 
-		var wrapper RISLiveMessage
+		var wrapper ristypes.RISLiveMessage
-		if err := json.Unmarshal(line, &wrapper); err != nil {
+
+		err := json.Unmarshal(line, &wrapper)
+		if err != nil {
 			t.Fatalf("unmarshal message %d: %v", messages+1, err)
 		}
+
 		messages++
 
 		msg := wrapper.Data
 		if msg.Community != nil {
-			t.Errorf("message %d: Community decoded, want empty: %v", messages, msg.Community)
+			t.Errorf("message %d: Community decoded, want empty: %v",
+				messages, msg.Community)
 		}
+
 		if msg.Raw != "" {
 			t.Errorf("message %d: Raw decoded, want empty", messages)
 		}
 
-		if len(msg.Path) > 0 {
+		if len(msg.Path.Sequence) > 0 {
 			withPath++
 		}
+
 		if len(msg.Announcements) > 0 {
 			withAnnouncements++
 		}
 	}
-	if err := scanner.Err(); err != nil {
+
+	err = scanner.Err()
+	if err != nil {
 		t.Fatalf("scan fixtures: %v", err)
 	}
 
@@ -62,6 +79,99 @@ func TestDecodeDropsCommunityAndRaw(t *testing.T) {
 	// The fixtures include announcement messages; a used field must still decode,
 	// otherwise an empty Community/Raw would prove nothing.
 	if withPath == 0 || withAnnouncements == 0 {
-		t.Fatalf("used fields did not decode: withPath=%d withAnnouncements=%d", withPath, withAnnouncements)
+		t.Fatalf("used fields did not decode: withPath=%d withAnnouncements=%d",
+			withPath, withAnnouncements)
+	}
+}
+
+// TestASPathKeepsATrailingSetApart decodes AS paths as the feed writes them, an
+// AS set as an array inside the path. A set of several members that ends the
+// path is kept apart from the sequence, and the path has no origin, since any
+// member may be it. A set of one member counts as that AS, and a set anywhere
+// else is flattened into the sequence.
+func TestASPathKeepsATrailingSetApart(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		path         string
+		wantSequence []int
+		wantSet      []int
+		wantOrigin   int
+	}{
+		{
+			path:         "[64496, 64497]",
+			wantSequence: []int{64496, 64497},
+			wantOrigin:   64497,
+		},
+		{
+			path:         "[64496, 64497, [64498, 64499]]",
+			wantSequence: []int{64496, 64497},
+			wantSet:      []int{64498, 64499},
+		},
+		{
+			path:         "[64496, 64497, [64498]]",
+			wantSequence: []int{64496, 64497, 64498},
+			wantOrigin:   64498,
+		},
+		{
+			path:         "[64496, [64497, 64498], 64499]",
+			wantSequence: []int{64496, 64497, 64498, 64499},
+			wantOrigin:   64499,
+		},
+		{path: "[]"},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.path, func(t *testing.T) {
+			t.Parallel()
+
+			var got ristypes.ASPath
+
+			err := json.Unmarshal([]byte(tt.path), &got)
+			if err != nil {
+				t.Fatalf("unmarshal %s: %v", tt.path, err)
+			}
+
+			if !slices.Equal(got.Sequence, tt.wantSequence) ||
+				!slices.Equal(got.Set, tt.wantSet) {
+				t.Errorf("%s decoded to sequence %v and set %v, want %v and %v",
+					tt.path, got.Sequence, got.Set, tt.wantSequence, tt.wantSet)
+			}
+
+			origin := got.Origin()
+			if origin != tt.wantOrigin {
+				t.Errorf("origin of %s = %d, want %d", tt.path, origin, tt.wantOrigin)
+			}
+		})
+	}
+}
+
+// TestASPathWithASZeroDecodesAsAnEmptyPath decodes AS paths that hold AS 0
+// before their end, at their end, or in an AS set. RFC 7607 makes such a path
+// malformed, so each decodes as an empty path.
+func TestASPathWithASZeroDecodesAsAnEmptyPath(t *testing.T) {
+	t.Parallel()
+
+	for _, path := range []string{
+		"[64496, 0, 64497]",
+		"[64496, 0]",
+		"[64496, [0, 64497], 64498]",
+		"[64496, [64497, 0]]",
+	} {
+		t.Run(path, func(t *testing.T) {
+			t.Parallel()
+
+			var got ristypes.ASPath
+
+			err := json.Unmarshal([]byte(path), &got)
+			if err != nil {
+				t.Fatalf("unmarshal %s: %v", path, err)
+			}
+
+			if len(got.Sequence) > 0 || len(got.Set) > 0 {
+				t.Errorf("%s decoded to sequence %v and set %v, want an empty path",
+					path, got.Sequence, got.Set)
+			}
+		})
 	}
 }

internal/routetable/benchmark_test.go

@@ -0,0 +1,155 @@
+package routetable_test
+
+import (
+	"fmt"
+	"math"
+	"math/rand/v2"
+	"net/netip"
+	"runtime"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+// Parameters for BenchmarkApplyMixedBatch.
+const (
+	// benchPeers is how many peers there are, about as many as RIS has.
+	benchPeers = 1400
+	// benchPrefixes is how many IPv4 /24 prefixes the table holds, and
+	// benchPeersPerPrefix how many peers, picked at random, carry each.
+	benchPrefixes       = 100000
+	benchPeersPerPrefix = 20
+	// benchBatchSize is how many updates the timed batch holds, as many as
+	// the prefix handler's largest batch (prefixBatchMaxSize in
+	// internal/routewatch).
+	benchBatchSize = 250000
+	// In the timed batch every benchWithdrawalEvery-th update is a
+	// withdrawal, and every benchNewOriginEvery-th announcement has an
+	// origin AS other than the prefix's.
+	benchWithdrawalEvery = 5
+	benchNewOriginEvery  = 7
+	// benchOrigins is how many origin ASes the prefixes are spread over,
+	// from benchFirstASN on: 20 prefixes each, about as many as an origin AS
+	// has on average in the full table.
+	benchOrigins  = 5000
+	benchFirstASN = 64500
+	// benchUpdatesPerSecond is how many updates share one time.
+	benchUpdatesPerSecond = 1000
+	benchMaskLength       = 24
+)
+
+// benchUpdate is one update of the benchmark.
+type benchUpdate struct {
+	prefix   netip.Prefix
+	origin   int
+	route    routetable.Route
+	withdraw bool
+}
+
+// BenchmarkApplyMixedBatch times one batch of 250,000 updates of routes picked
+// at random: announcements, withdrawals and announcements with a new origin AS.
+// The table it applies them to holds 100,000 IPv4 prefixes from 5,000 origin
+// ASes, each carried by 20 of 1,400 peers. It reports the heap bytes that table
+// takes per prefix and origin AS pair. It is run by hand, not by make test, for
+// example in the test image:
+//
+//	go test -run '^$' -bench MixedBatch -benchtime 10x ./internal/routetable/
+func BenchmarkApplyMixedBatch(b *testing.B) {
+	//nolint:gosec // The benchmark needs the same updates every run.
+	rng := rand.New(rand.NewPCG(1, 2))
+	peers := make([]string, benchPeers)
+
+	for i := range peers {
+		peers[i] = fmt.Sprintf("198.18.%d.%d", i/256, i%256)
+	}
+
+	before := heapBytes()
+	table := routetable.New(routetable.Params{})
+	updates := 0
+
+	for prefix := range benchPrefixes {
+		for range benchPeersPerPrefix {
+			peer := peers[rng.IntN(benchPeers)]
+			apply(table, benchAnnouncement(rng, prefix, peer, updates, false))
+			updates++
+		}
+	}
+
+	// The record of changes is handed over every second while routewatch
+	// runs, so it is not part of the table's size.
+	table.TakeChanges(math.MaxInt)
+
+	bytesPerPair := float64(heapBytes()-before) / benchPrefixes
+
+	batch := make([]benchUpdate, 0, benchBatchSize)
+	for i := range benchBatchSize {
+		update := benchAnnouncement(rng, rng.IntN(benchPrefixes),
+			peers[rng.IntN(benchPeers)], updates+i, i%benchNewOriginEvery == 0)
+		update.withdraw = i%benchWithdrawalEvery == 0
+		batch = append(batch, update)
+	}
+
+	for b.Loop() {
+		for _, update := range batch {
+			apply(table, update)
+		}
+	}
+
+	// After the loop: b.Loop drops the metrics reported before it starts.
+	b.ReportMetric(bytesPerPair, "bytes/pair")
+}
+
+// benchAnnouncement returns the announcement of the prefix numbered prefix, a
+// /24 from 1.0.0.0/24 on, by peer. It is the updates-th update of the
+// benchmark, and its AS path holds four ASes from the peer's AS to the origin
+// AS. newOrigin gives it an origin AS other than the prefix's.
+func benchAnnouncement(
+	rng *rand.Rand, prefix int, peer string, updates int, newOrigin bool,
+) benchUpdate {
+	origin := benchFirstASN + prefix%benchOrigins
+	if newOrigin {
+		origin += benchOrigins
+	}
+
+	start := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
+
+	return benchUpdate{
+		prefix: netip.MustParsePrefix(fmt.Sprintf("%d.%d.%d.0/%d",
+			1+prefix>>16, (prefix>>8)%256, prefix%256, benchMaskLength)),
+		origin: origin,
+		route: routetable.Route{
+			Peer: peer,
+			ASPath: []int{
+				benchFirstASN + rng.IntN(benchOrigins),
+				benchFirstASN + rng.IntN(benchOrigins),
+				benchFirstASN + rng.IntN(benchOrigins),
+				origin,
+			},
+			Time: start.Add(time.Duration(updates/benchUpdatesPerSecond) *
+				time.Second),
+		},
+	}
+}
+
+// apply applies update to table.
+func apply(table *routetable.Table, update benchUpdate) {
+	if update.withdraw {
+		table.Withdraw(update.prefix, update.route.Peer)
+
+		return
+	}
+
+	table.Announce(update.prefix, update.origin, update.route)
+}
+
+// heapBytes returns the bytes of the live heap objects, after a collection.
+func heapBytes() uint64 {
+	runtime.GC()
+
+	var stats runtime.MemStats
+
+	runtime.ReadMemStats(&stats)
+
+	return stats.HeapAlloc
+}

internal/routetable/reads.go

@@ -0,0 +1,233 @@
+package routetable
+
+import (
+	"net/netip"
+	"slices"
+	"time"
+
+	"github.com/bits-and-blooms/bitset"
+)
+
+// PrefixSummary is what the table answers about one prefix.
+type PrefixSummary struct {
+	Prefix netip.Prefix
+	// OriginASN is the origin AS that the most peers carry the prefix with.
+	// When several have as many peers, it is the lowest AS number.
+	OriginASN int
+	// Peers is how many peers carry the prefix.
+	Peers int
+	// Newest is the time of the newest kept route of the prefix, or nil when
+	// it keeps none.
+	Newest *time.Time
+}
+
+// OriginatedPrefix is one prefix an AS originates.
+type OriginatedPrefix struct {
+	Prefix netip.Prefix
+	// Newest is the time of the newest kept route of the prefix with that
+	// origin AS, or nil when it keeps none.
+	Newest *time.Time
+}
+
+// Counts are the table's figures for the stats.
+type Counts struct {
+	// Pairs counts the prefix and origin AS pairs that some peer carries.
+	Pairs int
+	// Peers counts the peers that have a number: those given in
+	// Params.PeerNumbers and those numbered since.
+	Peers        int
+	IPv4Prefixes int
+	IPv6Prefixes int
+	// IPv4PrefixesByLength[n] and IPv6PrefixesByLength[n] are how many
+	// prefixes of mask length n some peer carries.
+	IPv4PrefixesByLength []int
+	IPv6PrefixesByLength []int
+	// IPv4Routes and IPv6Routes count peer and prefix pairs: each peer that
+	// carries a prefix counts once for it.
+	IPv4Routes int
+	IPv6Routes int
+	// OriginsWithoutKeptRoute counts the prefix and origin AS pairs that peers
+	// carry but whose kept routes were all withdrawn.
+	OriginsWithoutKeptRoute int
+	// OldestKeptRoute and NewestKeptRoute are the times of the oldest and the
+	// newest kept routes, or nil when the table keeps none.
+	OldestKeptRoute *time.Time
+	NewestKeptRoute *time.Time
+}
+
+// Lookup returns the most specific prefix that holds addr and that some peer
+// carries. It ignores prefixes shorter than /8 for IPv4 and /16 for IPv6. It
+// makes one map lookup per mask length, and reports false when no prefix
+// holds addr.
+func (t *Table) Lookup(addr netip.Addr) (PrefixSummary, bool) {
+	t.mu.RLock()
+	defer t.mu.RUnlock()
+
+	// An IPv4 address written in IPv6 form (::ffff:192.0.2.1) is looked up as
+	// IPv4.
+	addr = addr.Unmap()
+	f := t.familyOf(addr)
+
+	for bits := addr.BitLen(); bits >= f.minLookupLength; bits-- {
+		prefix := netip.PrefixFrom(addr, bits).Masked()
+
+		origins, found := f.byLength[bits][prefix]
+		if found {
+			return summarize(prefix, origins), true
+		}
+	}
+
+	return PrefixSummary{}, false
+}
+
+// OriginatedPrefixes returns the prefixes AS asn originates, in prefix order.
+func (t *Table) OriginatedPrefixes(asn int) []OriginatedPrefix {
+	t.mu.RLock()
+	defer t.mu.RUnlock()
+
+	prefixes := make([]OriginatedPrefix, 0, len(t.byOrigin[asn]))
+	for prefix, o := range t.byOrigin[asn] {
+		prefixes = append(prefixes,
+			OriginatedPrefix{Prefix: prefix, Newest: o.newestTime()})
+	}
+
+	slices.SortFunc(prefixes, func(a, b OriginatedPrefix) int {
+		return a.Prefix.Compare(b.Prefix)
+	})
+
+	return prefixes
+}
+
+// Origins returns each origin AS of prefix, in AS number order, with its peers
+// and its kept routes. It returns nil when no peer carries the prefix.
+func (t *Table) Origins(prefix netip.Prefix) []Origin {
+	t.mu.RLock()
+	defer t.mu.RUnlock()
+
+	if !prefix.IsValid() {
+		return nil
+	}
+
+	var origins []Origin
+
+	for _, o := range t.origins(prefix.Masked()) {
+		origins = append(origins, Origin{
+			ASN:    o.asn,
+			Peers:  t.addresses(&o.peers),
+			Routes: t.exportRoutes(o.routes),
+		})
+	}
+
+	return origins
+}
+
+// SamplePrefixes returns up to limit of the prefixes of IP version ipVersion
+// (4 or 6) and mask length maskLength. It reads only the prefixes it returns.
+// Go starts each walk over a map at a random place, so each call returns
+// another sample.
+func (t *Table) SamplePrefixes(ipVersion, maskLength, limit int) []PrefixSummary {
+	t.mu.RLock()
+	defer t.mu.RUnlock()
+
+	var f *family
+
+	switch ipVersion {
+	case ipVersion4:
+		f = &t.ipv4
+	case ipVersion6:
+		f = &t.ipv6
+	default:
+		return nil
+	}
+
+	if maskLength < 0 || maskLength >= len(f.byLength) || limit <= 0 {
+		return nil
+	}
+
+	prefixes := f.byLength[maskLength]
+	sample := make([]PrefixSummary, 0, min(limit, len(prefixes)))
+
+	for prefix, origins := range prefixes {
+		if len(sample) == limit {
+			break
+		}
+
+		sample = append(sample, summarize(prefix, origins))
+	}
+
+	return sample
+}
+
+// Counts returns the table's figures for the stats. The writes keep them up to
+// date, so this reads no prefix.
+func (t *Table) Counts() Counts {
+	t.mu.RLock()
+	defer t.mu.RUnlock()
+
+	counts := Counts{
+		Peers:                   len(t.peerNumbers),
+		IPv4PrefixesByLength:    t.ipv4.prefixesByLength(),
+		IPv6PrefixesByLength:    t.ipv6.prefixesByLength(),
+		IPv4Routes:              t.ipv4.routes,
+		IPv6Routes:              t.ipv6.routes,
+		OriginsWithoutKeptRoute: t.originsWithoutKeptRoute,
+	}
+
+	for _, n := range counts.IPv4PrefixesByLength {
+		counts.IPv4Prefixes += n
+	}
+
+	for _, n := range counts.IPv6PrefixesByLength {
+		counts.IPv6Prefixes += n
+	}
+
+	for _, prefixes := range t.byOrigin {
+		counts.Pairs += len(prefixes)
+	}
+
+	oldest, found := t.keptTimes.Min()
+	if found {
+		counts.OldestKeptRoute = &oldest.time
+	}
+
+	newest, found := t.keptTimes.Max()
+	if found {
+		counts.NewestKeptRoute = &newest.time
+	}
+
+	return counts
+}
+
+// addresses returns the addresses of the peers in peers, in number order.
+func (t *Table) addresses(peers *bitset.BitSet) []string {
+	addresses := make([]string, 0, peers.Count())
+	for number := range peers.EachSet() {
+		addresses = append(addresses, t.peerAddresses[number])
+	}
+
+	return addresses
+}
+
+// summarize returns the summary of prefix, whose origins are in AS number
+// order.
+func summarize(prefix netip.Prefix, origins []*origin) PrefixSummary {
+	summary := PrefixSummary{Prefix: prefix}
+	mostPeers := 0
+
+	for _, o := range origins {
+		peers := int(o.peers.Count())
+		summary.Peers += peers
+
+		if peers > mostPeers {
+			mostPeers = peers
+			summary.OriginASN = o.asn
+		}
+
+		newest := o.newestTime()
+		if newest != nil && (summary.Newest == nil || newest.After(*summary.Newest)) {
+			summary.Newest = newest
+		}
+	}
+
+	return summary
+}

internal/routetable/reads_test.go

@@ -0,0 +1,265 @@
+package routetable_test
+
+import (
+	"net/netip"
+	"reflect"
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+// announceAll has each listed peer announce prefix with origin AS asn, the
+// peer at index i i*10 seconds after start.
+func announceAll(
+	table *routetable.Table, prefix string, asn, start int, peers ...string,
+) {
+	const secondsBetweenPeers = 10
+
+	for i, peer := range peers {
+		table.Announce(netip.MustParsePrefix(prefix), asn,
+			routeOf(peer, asn, start+i*secondsBetweenPeers))
+	}
+}
+
+func TestLookup(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+	announceAll(table, "10.0.0.0/8", asA, 10, peerA)
+	// Three peers with two origins: asB has more of them.
+	announceAll(table, "10.1.0.0/16", asA, 10, peerA)
+	announceAll(table, "10.1.0.0/16", asB, 20, peerB, peerC)
+	// Two origins with one peer each: the lower AS number wins.
+	announceAll(table, prefixV4, asB, 20, peerB)
+	announceAll(table, prefixV4, asA, 10, peerA)
+	// An origin whose kept routes were withdrawn: no newest time.
+	announceAll(table, "10.9.0.0/16", asA, 10, peerA, peerB, peerC)
+	table.Withdraw(netip.MustParsePrefix("10.9.0.0/16"), peerB)
+	table.Withdraw(netip.MustParsePrefix("10.9.0.0/16"), peerC)
+	announceAll(table, "2001:db8::/32", asA, 10, peerA)
+	announceAll(table, prefixV6, asB, 20, peerB)
+	announceAll(table, "3000::/16", asA, 10, peerA)
+	// Shorter than the lookup uses.
+	announceAll(table, "192.0.0.0/7", asA, 10, peerA)
+	announceAll(table, "2000::/15", asA, 10, peerA)
+
+	tests := []struct {
+		addr string
+		// want is the summary's prefix, or empty when none is found.
+		want      string
+		originASN int
+		peers     int
+		newest    int // in seconds after the start; 0 for none
+	}{
+		{"10.1.2.3", prefixV4, asA, 2, 20},
+		{"::ffff:10.1.2.3", prefixV4, asA, 2, 20},
+		{"10.1.3.4", "10.1.0.0/16", asB, 3, 30},
+		{"10.2.0.1", "10.0.0.0/8", asA, 1, 10},
+		{"10.9.0.1", "10.9.0.0/16", asA, 1, 0},
+		{"192.0.2.1", "", 0, 0, 0},
+		{"2001:db8:1::5", prefixV6, asB, 1, 20},
+		{"2001:db8:2::1", "2001:db8::/32", asA, 1, 10},
+		{"3000::1", "3000::/16", asA, 1, 10},
+		{"2001:db9::1", "", 0, 0, 0},
+	}
+
+	for _, tc := range tests {
+		t.Run(tc.addr, func(t *testing.T) {
+			t.Parallel()
+
+			got, found := table.Lookup(netip.MustParseAddr(tc.addr))
+			if tc.want == "" {
+				if found {
+					t.Errorf("Lookup(%s) = %+v, want none", tc.addr, got)
+				}
+
+				return
+			}
+
+			want := routetable.PrefixSummary{
+				Prefix:    netip.MustParsePrefix(tc.want),
+				OriginASN: tc.originASN,
+				Peers:     tc.peers,
+			}
+			if tc.newest != 0 {
+				want.Newest = timeAt(tc.newest)
+			}
+
+			if !found || !reflect.DeepEqual(got, want) {
+				t.Errorf("Lookup(%s) = %+v, %v; want %+v", tc.addr, got, found, want)
+			}
+		})
+	}
+}
+
+func TestOriginatedPrefixes(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+	announceAll(table, "2001:db8::/32", asA, 30, peerA)
+	announceAll(table, prefixV4, asA, 10, peerA)
+	// asB's newer route for the same prefix does not count for asA.
+	announceAll(table, prefixV4, asB, 40, peerB)
+	announceAll(table, "10.1.0.0/16", asA, 10, peerA, peerB)
+	announceAll(table, "10.9.0.0/16", asA, 10, peerA, peerB, peerC)
+	table.Withdraw(netip.MustParsePrefix("10.9.0.0/16"), peerB)
+	table.Withdraw(netip.MustParsePrefix("10.9.0.0/16"), peerC)
+
+	want := []routetable.OriginatedPrefix{
+		{Prefix: netip.MustParsePrefix("10.1.0.0/16"), Newest: timeAt(20)},
+		{Prefix: netip.MustParsePrefix(prefixV4), Newest: timeAt(10)},
+		{Prefix: netip.MustParsePrefix("10.9.0.0/16")},
+		{Prefix: netip.MustParsePrefix("2001:db8::/32"), Newest: timeAt(30)},
+	}
+
+	got := table.OriginatedPrefixes(asA)
+	if !reflect.DeepEqual(got, want) {
+		t.Errorf("OriginatedPrefixes(%d) = %+v, want %+v", asA, got, want)
+	}
+
+	got = table.OriginatedPrefixes(asC)
+	if len(got) != 0 {
+		t.Errorf("OriginatedPrefixes(%d) = %+v, want none", asC, got)
+	}
+}
+
+func TestSamplePrefixes(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+
+	announced := []string{
+		"10.0.0.0/24", "10.0.1.0/24", "10.0.2.0/24", "10.0.3.0/24",
+		"10.0.4.0/24", "10.0.0.0/16",
+		"2001:db8::/48", "2001:db8:1::/48", "2001:db8:2::/48",
+	}
+	for _, prefix := range announced {
+		announceAll(table, prefix, asA, 10, peerA)
+	}
+
+	tests := []struct {
+		name                         string
+		ipVersion, maskLength, limit int
+		want                         int
+	}{
+		{"fewer than there are", 4, 24, 3, 3},
+		{"more than there are", 4, 24, 10, 5},
+		{"IPv6", 6, 48, 10, 3},
+		{"a mask length with none", 4, 20, 10, 0},
+		{"no such IP version", 5, 24, 10, 0},
+		{"a mask length too long", 4, 33, 10, 0},
+		{"a negative mask length", 4, -1, 10, 0},
+		{"a limit of zero", 4, 24, 0, 0},
+	}
+
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+
+			sample := table.SamplePrefixes(tc.ipVersion, tc.maskLength, tc.limit)
+			if len(sample) != tc.want {
+				t.Fatalf("SamplePrefixes(%d, %d, %d) returned %d prefixes, want %d",
+					tc.ipVersion, tc.maskLength, tc.limit, len(sample), tc.want)
+			}
+
+			seen := make(map[netip.Prefix]bool)
+
+			for _, got := range sample {
+				want := routetable.PrefixSummary{
+					Prefix: got.Prefix, OriginASN: asA, Peers: 1, Newest: timeAt(10),
+				}
+				if seen[got.Prefix] || got.Prefix.Bits() != tc.maskLength ||
+					!reflect.DeepEqual(got, want) {
+					t.Errorf("SamplePrefixes(%d, %d, %d) returned %+v",
+						tc.ipVersion, tc.maskLength, tc.limit, got)
+				}
+
+				seen[got.Prefix] = true
+			}
+		})
+	}
+}
+
+func TestCounts(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+	checkCounts(t, table, routetable.Counts{
+		IPv4PrefixesByLength: byLength(ipv4Bits, nil),
+		IPv6PrefixesByLength: byLength(ipv6Bits, nil),
+	})
+
+	announceAll(table, prefixV4, asA, 10, peerA, peerB)
+	announceAll(table, "10.2.0.0/16", asA, 15, peerA)
+	announceAll(table, "2001:db8::/32", asA, 5, peerA)
+	announceAll(table, "2001:db8::/32", asA, 25, peerC)
+	checkCounts(t, table, routetable.Counts{
+		Pairs:                3,
+		Peers:                3,
+		IPv4Prefixes:         2,
+		IPv6Prefixes:         1,
+		IPv4PrefixesByLength: byLength(ipv4Bits, map[int]int{16: 1, 24: 1}),
+		IPv6PrefixesByLength: byLength(ipv6Bits, map[int]int{32: 1}),
+		IPv4Routes:           3,
+		IPv6Routes:           2,
+		OldestKeptRoute:      timeAt(5),
+		NewestKeptRoute:      timeAt(25),
+	})
+
+	// peerC's route drops peerA's from the kept ones, and peerA leaves the
+	// IPv6 prefix.
+	announceAll(table, prefixV4, asA, 30, peerC)
+	table.Withdraw(netip.MustParsePrefix("2001:db8::/32"), peerA)
+	checkCounts(t, table, routetable.Counts{
+		Pairs:                3,
+		Peers:                3,
+		IPv4Prefixes:         2,
+		IPv6Prefixes:         1,
+		IPv4PrefixesByLength: byLength(ipv4Bits, map[int]int{16: 1, 24: 1}),
+		IPv6PrefixesByLength: byLength(ipv6Bits, map[int]int{32: 1}),
+		IPv4Routes:           4,
+		IPv6Routes:           1,
+		OldestKeptRoute:      timeAt(15),
+		NewestKeptRoute:      timeAt(30),
+	})
+
+	// peerC was the last peer of the IPv6 prefix. It keeps its number.
+	table.PeerDown(peerC)
+	checkCounts(t, table, routetable.Counts{
+		Pairs:                2,
+		Peers:                3,
+		IPv4Prefixes:         2,
+		IPv4PrefixesByLength: byLength(ipv4Bits, map[int]int{16: 1, 24: 1}),
+		IPv6PrefixesByLength: byLength(ipv6Bits, nil),
+		IPv4Routes:           3,
+		OldestKeptRoute:      timeAt(15),
+		NewestKeptRoute:      timeAt(20),
+	})
+}
+
+// The address lengths of the two IP versions.
+const (
+	ipv4Bits = 32
+	ipv6Bits = 128
+)
+
+// byLength returns the prefix counts by mask length, for addresses of bits
+// bits, with counts[n] prefixes of mask length n and none of any other.
+func byLength(bits int, counts map[int]int) []int {
+	all := make([]int, bits+1)
+	for n, count := range counts {
+		all[n] = count
+	}
+
+	return all
+}
+
+// checkCounts checks that the table's counts are want.
+func checkCounts(t *testing.T, table *routetable.Table, want routetable.Counts) {
+	t.Helper()
+
+	got := table.Counts()
+	if !reflect.DeepEqual(got, want) {
+		t.Errorf("Counts() = %+v, want %+v", got, want)
+	}
+}

internal/routetable/routetable.go

@@ -0,0 +1,677 @@
+// Package routetable holds the whole routing table in memory. For each prefix
+// and origin AS, it keeps the peers whose current route to the prefix has that
+// origin AS. It also keeps the two newest of those routes in full. It applies
+// announcements, withdrawals and peers' sessions going down. It answers the
+// route reads that the pages and the API need.
+package routetable
+
+import (
+	"cmp"
+	"maps"
+	"net/netip"
+	"slices"
+	"sync"
+	"time"
+
+	"github.com/bits-and-blooms/bitset"
+	"github.com/google/btree"
+)
+
+const (
+	// keptRoutesPerOrigin is how many routes of each prefix and origin AS are
+	// kept in full: the newest ones. Every other peer that carries the prefix
+	// with that origin AS keeps only its bit in the peer set.
+	keptRoutesPerOrigin = 2
+
+	// keptTimesDegree is the degree of the B-tree that holds the times of the
+	// kept routes in order: each of its nodes holds up to 63 times.
+	keptTimesDegree = 32
+
+	ipv4Bits = 32
+	ipv6Bits = 128
+
+	// The address lookup ignores prefixes shorter than these mask lengths. A
+	// default route or a route this short only says where a peer sends traffic
+	// it has no better route for; it does not say who holds the address.
+	minLookupMaskLengthV4 = 8
+	minLookupMaskLengthV6 = 16
+
+	// The IP versions SamplePrefixes takes.
+	ipVersion4 = 4
+	ipVersion6 = 6
+)
+
+// Route is one peer's route to one prefix. The next hop is the peer itself, so
+// it is not stored.
+type Route struct {
+	// Peer is the peer's address.
+	Peer   string
+	ASPath []int
+	// Time is when the peer announced the route.
+	Time time.Time
+}
+
+// Origin is one origin AS of a prefix: the peers whose current route to the
+// prefix has that origin AS, and the newest of those routes in full.
+type Origin struct {
+	// ASN is the origin AS. 0 means no known origin: the path ends in an AS
+	// set of several members.
+	ASN int
+	// Peers are the peers' addresses, in the order of their numbers: the
+	// order they were first seen in.
+	Peers []string
+	// Routes are the kept routes, newest first: at most two, and none when
+	// both were withdrawn while other peers still carry the prefix.
+	Routes []Route
+}
+
+// Change is the current content of one prefix and origin AS pair that changed:
+// what a database row for the pair holds.
+type Change struct {
+	Prefix    netip.Prefix
+	OriginASN int
+	// Removed says that no peer carries the prefix with this origin AS any
+	// more. Peers and Routes are then nil.
+	Removed bool
+	// Peers holds the number of each peer that carries the prefix with this
+	// origin AS.
+	Peers *bitset.BitSet
+	// Routes are the kept routes, newest first.
+	Routes []Route
+}
+
+// Params configures New.
+type Params struct {
+	// PeerNumbers holds the number each peer was given before, by address. A
+	// peer not in it gets the next number above all of them the first time it
+	// announces a route.
+	PeerNumbers map[string]uint
+}
+
+// Table is the routing table. Its methods are safe for concurrent use.
+type Table struct {
+	mu sync.RWMutex
+
+	ipv4 family
+	ipv6 family
+	// byOrigin holds, for each origin AS, the prefixes it originates.
+	byOrigin map[int]map[netip.Prefix]*origin
+
+	// peerNumbers and peerAddresses give each peer's number by its address
+	// and its address by its number. nextPeerNumber is the number the next
+	// new peer gets.
+	peerNumbers    map[string]uint
+	peerAddresses  map[uint]string
+	nextPeerNumber uint
+	// peerPrefixes[n] counts the prefixes the peer numbered n carries, for
+	// each n below nextPeerNumber. It is a slice, not a map, because Load
+	// counts once for every peer of every pair it loads: with a map, reading
+	// a full table at startup took several times as long.
+	peerPrefixes []int
+
+	// newPeers and changed are what TakeChanges hands over: the peers
+	// numbered since it last ran, and the prefix and origin AS pairs changed
+	// since it last took them.
+	newPeers map[string]uint
+	changed  map[pair]struct{}
+
+	// keptTimes counts the kept routes at each time, in time order, so its
+	// first and last items are the oldest and newest kept routes.
+	keptTimes *btree.BTreeG[keptTime]
+	// originsWithoutKeptRoute counts the origins that peers carry but that
+	// keep no route.
+	originsWithoutKeptRoute int
+}
+
+// New returns an empty table that numbers new peers on from
+// params.PeerNumbers.
+func New(params Params) *Table {
+	t := &Table{
+		ipv4:          newFamily(ipv4Bits, minLookupMaskLengthV4),
+		ipv6:          newFamily(ipv6Bits, minLookupMaskLengthV6),
+		byOrigin:      make(map[int]map[netip.Prefix]*origin),
+		peerNumbers:   make(map[string]uint, len(params.PeerNumbers)),
+		peerAddresses: make(map[uint]string, len(params.PeerNumbers)),
+		newPeers:      make(map[string]uint),
+		changed:       make(map[pair]struct{}),
+		keptTimes: btree.NewG(keptTimesDegree, func(a, b keptTime) bool {
+			return a.time.Before(b.time)
+		}),
+	}
+
+	for address, number := range params.PeerNumbers {
+		t.peerNumbers[address] = number
+		t.peerAddresses[number] = address
+		t.nextPeerNumber = max(t.nextPeerNumber, number+1)
+	}
+
+	t.peerPrefixes = make([]int, t.nextPeerNumber)
+
+	return t
+}
+
+// Announce applies an announcement: the peer's current route to prefix is now
+// r, with origin AS originASN. A peer that carried the prefix with another
+// origin AS leaves that origin's peer set, and its route there is dropped. r
+// is kept in full when it is one of the two newest routes of its prefix and
+// origin AS; otherwise only the peer's bit is.
+//
+// An announcement whose path is empty or holds AS 0 is a withdrawal, and the
+// caller passes it to Withdraw instead.
+func (t *Table) Announce(prefix netip.Prefix, originASN int, r Route) {
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	if !prefix.IsValid() {
+		return
+	}
+
+	prefix = prefix.Masked()
+	peer := t.numberPeer(r.Peer)
+
+	current := originWithPeer(t.origins(prefix), peer)
+	if current != nil && current.asn != originASN {
+		t.leave(prefix, current, peer)
+	}
+
+	t.join(prefix, t.findOrAddOrigin(prefix, originASN), route{
+		peer: peer,
+		path: slices.Clone(r.ASPath),
+		time: r.Time,
+	})
+}
+
+// Withdraw applies a withdrawal: the peer no longer carries prefix. It leaves
+// the prefix's peer set, and its kept route, if any, is dropped.
+func (t *Table) Withdraw(prefix netip.Prefix, peer string) {
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	number, known := t.peerNumbers[peer]
+	if !known || !prefix.IsValid() {
+		return
+	}
+
+	prefix = prefix.Masked()
+
+	o := originWithPeer(t.origins(prefix), number)
+	if o != nil {
+		t.leave(prefix, o, number)
+	}
+}
+
+// PeerDown applies a peer's session going down: the peer leaves every peer
+// set, and every route of it is dropped. When the peer carries any prefix, it
+// visits every prefix in the table.
+func (t *Table) PeerDown(peer string) {
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	number, known := t.peerNumbers[peer]
+	if !known || t.peerPrefixes[number] == 0 {
+		return
+	}
+
+	for _, f := range []*family{&t.ipv4, &t.ipv6} {
+		for _, prefixes := range f.byLength {
+			for prefix, origins := range prefixes {
+				o := originWithPeer(origins, number)
+				if o != nil {
+					t.leave(prefix, o, number)
+				}
+			}
+		}
+	}
+}
+
+// Load puts one prefix and origin AS pair that the database holds into the
+// table, as startup does with each row it reads. The change's peers join the
+// pair's peer set, and its routes are kept as an announcement's would be.
+// Unlike an update, Load records no change, so TakeChanges does not hand the
+// pair back to be written again. Loading a pair twice changes nothing.
+//
+// The caller makes sure that every peer change names, in its peer set or in
+// its routes, has a number in Params.PeerNumbers.
+func (t *Table) Load(change Change) {
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	// A removed pair has no peers, and a pair without peers is not in the
+	// table.
+	if change.Peers == nil || change.Peers.None() || !change.Prefix.IsValid() {
+		return
+	}
+
+	prefix := change.Prefix.Masked()
+	o := t.findOrAddOrigin(prefix, change.OriginASN)
+	wasWithoutKeptRoute := o.withoutKeptRoute()
+
+	// The peers not yet in the peer set join it in one union, which takes
+	// less time than setting their bits one at a time.
+	joining := change.Peers.Difference(&o.peers)
+	o.peers.InPlaceUnion(joining)
+	t.familyOf(prefix.Addr()).routes += int(joining.Count())
+
+	for peer := range joining.EachSet() {
+		t.peerPrefixes[peer]++
+	}
+
+	// Oldest first, as they were announced, so that two routes with the same
+	// time are kept in the order they had.
+	for _, r := range slices.Backward(change.Routes) {
+		peer := t.peerNumbers[r.Peer]
+
+		t.dropKeptRoute(o, peer)
+		t.keepIfNewest(o, route{
+			peer: peer,
+			path: slices.Clone(r.ASPath),
+			time: r.Time,
+		})
+	}
+
+	t.countOriginWithoutKeptRoute(o, wasWithoutKeptRoute)
+}
+
+// TakeChanges returns at most maxChanges of the prefix and origin AS pairs that
+// changed since they were last taken, in prefix order, each with its current
+// content. The other changed pairs wait for a later call. It also returns every
+// peer numbered since it last ran, by address, so a pair taken later names only
+// peers already returned. It then forgets what it returned.
+func (t *Table) TakeChanges(maxChanges int) ([]Change, map[string]uint) {
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	changes := make([]Change, 0, min(maxChanges, len(t.changed)))
+	for key := range t.changed {
+		if len(changes) == maxChanges {
+			break
+		}
+
+		changes = append(changes, t.currentContent(key))
+		delete(t.changed, key)
+	}
+
+	// A map keeps the memory of its largest size, so an emptied one is
+	// replaced.
+	if len(t.changed) == 0 {
+		t.changed = make(map[pair]struct{})
+	}
+
+	slices.SortFunc(changes, func(a, b Change) int {
+		return cmp.Or(a.Prefix.Compare(b.Prefix),
+			cmp.Compare(a.OriginASN, b.OriginASN))
+	})
+
+	newPeers := t.newPeers
+	t.newPeers = make(map[string]uint)
+
+	return changes, newPeers
+}
+
+// PutBackChanges records again what one TakeChanges call returned, for a
+// caller that could not write it. The next TakeChanges returns those peers
+// again, and those pairs wait to be taken again like any changed pair. Each is
+// then taken with its content at that time, so a later change of a pair
+// replaces the earlier.
+func (t *Table) PutBackChanges(changes []Change, newPeers map[string]uint) {
+	t.mu.Lock()
+	defer t.mu.Unlock()
+
+	for _, change := range changes {
+		t.changed[pair{prefix: change.Prefix, asn: change.OriginASN}] = struct{}{}
+	}
+
+	maps.Copy(t.newPeers, newPeers)
+}
+
+// ChangeCount returns how many changed prefix and origin AS pairs wait for
+// TakeChanges.
+func (t *Table) ChangeCount() int {
+	t.mu.RLock()
+	defer t.mu.RUnlock()
+
+	return len(t.changed)
+}
+
+// numberPeer returns the peer's number, and gives it the next one the first
+// time the table sees it.
+func (t *Table) numberPeer(address string) uint {
+	number, known := t.peerNumbers[address]
+	if known {
+		return number
+	}
+
+	number = t.nextPeerNumber
+	t.nextPeerNumber++
+	t.peerNumbers[address] = number
+	t.peerAddresses[number] = address
+	t.peerPrefixes = append(t.peerPrefixes, 0)
+	t.newPeers[address] = number
+
+	return number
+}
+
+// familyOf returns the part of the table for addr's IP version.
+func (t *Table) familyOf(addr netip.Addr) *family {
+	if addr.Is4() {
+		return &t.ipv4
+	}
+
+	return &t.ipv6
+}
+
+// prefixesOfLength returns the prefixes of the IP version and the mask length
+// of prefix, a valid masked prefix.
+func (t *Table) prefixesOfLength(prefix netip.Prefix) map[netip.Prefix][]*origin {
+	return t.familyOf(prefix.Addr()).byLength[prefix.Bits()]
+}
+
+// origins returns the origins of prefix, a valid masked prefix.
+func (t *Table) origins(prefix netip.Prefix) []*origin {
+	return t.prefixesOfLength(prefix)[prefix]
+}
+
+// findOrAddOrigin returns the origin of prefix with AS asn, added with no
+// peers when the prefix has none with that AS.
+func (t *Table) findOrAddOrigin(prefix netip.Prefix, asn int) *origin {
+	prefixes := t.prefixesOfLength(prefix)
+	origins := prefixes[prefix]
+
+	i, found := slices.BinarySearchFunc(origins, asn,
+		func(o *origin, asn int) int { return cmp.Compare(o.asn, asn) })
+	if found {
+		return origins[i]
+	}
+
+	o := &origin{asn: asn}
+	prefixes[prefix] = slices.Insert(origins, i, o)
+
+	if t.byOrigin[asn] == nil {
+		t.byOrigin[asn] = make(map[netip.Prefix]*origin)
+	}
+
+	t.byOrigin[asn][prefix] = o
+
+	return o
+}
+
+// removeOrigin removes o, which no peer carries any more, from prefix, and the
+// prefix from the table when it has no origin left.
+func (t *Table) removeOrigin(prefix netip.Prefix, o *origin) {
+	prefixes := t.prefixesOfLength(prefix)
+
+	origins := slices.DeleteFunc(prefixes[prefix],
+		func(other *origin) bool { return other == o })
+	if len(origins) == 0 {
+		delete(prefixes, prefix)
+	} else {
+		prefixes[prefix] = origins
+	}
+
+	delete(t.byOrigin[o.asn], prefix)
+
+	if len(t.byOrigin[o.asn]) == 0 {
+		delete(t.byOrigin, o.asn)
+	}
+}
+
+// join puts the peer of r in the peer set of o, an origin of prefix, and keeps
+// r in full when it is one of the two newest routes of o. The peer's earlier
+// route, if o keeps one, is replaced: it is no longer the peer's current route.
+// The pair is recorded as changed only when its peer set or kept routes change.
+func (t *Table) join(prefix netip.Prefix, o *origin, r route) {
+	if slices.ContainsFunc(o.routes, r.equal) {
+		// The peer announced its kept route again.
+		return
+	}
+
+	wasWithoutKeptRoute := o.withoutKeptRoute()
+
+	joined := !o.peers.Test(r.peer)
+	if joined {
+		o.peers.Set(r.peer)
+
+		t.familyOf(prefix.Addr()).routes++
+		t.peerPrefixes[r.peer]++
+	}
+
+	dropped := t.dropKeptRoute(o, r.peer)
+	kept := t.keepIfNewest(o, r)
+
+	if joined || dropped || kept {
+		t.recordChange(prefix, o, wasWithoutKeptRoute)
+	}
+}
+
+// leave takes peer out of the peer set of o, an origin of prefix, and drops
+// its kept route, if any.
+func (t *Table) leave(prefix netip.Prefix, o *origin, peer uint) {
+	wasWithoutKeptRoute := o.withoutKeptRoute()
+
+	o.peers.Clear(peer)
+	t.dropKeptRoute(o, peer)
+
+	t.familyOf(prefix.Addr()).routes--
+	t.peerPrefixes[peer]--
+
+	t.recordChange(prefix, o, wasWithoutKeptRoute)
+}
+
+// keepIfNewest keeps r in o unless both kept routes of o are newer, and says
+// whether it kept r. When o already keeps two routes, the older one is
+// dropped, and its peer keeps only its bit.
+func (t *Table) keepIfNewest(o *origin, r route) bool {
+	// The kept routes stay newest first: r goes before the first one that is
+	// not newer than it.
+	i := 0
+	for i < len(o.routes) && o.routes[i].time.After(r.time) {
+		i++
+	}
+
+	if i == keptRoutesPerOrigin {
+		return false
+	}
+
+	if len(o.routes) == keptRoutesPerOrigin {
+		t.dropKeptRoute(o, o.routes[keptRoutesPerOrigin-1].peer)
+	}
+
+	o.routes = slices.Insert(o.routes, i, r)
+	t.countKeptTime(r.time, 1)
+
+	return true
+}
+
+// dropKeptRoute drops the peer's route from the kept routes of o, if o keeps
+// one, and says whether it did.
+func (t *Table) dropKeptRoute(o *origin, peer uint) bool {
+	i := slices.IndexFunc(o.routes, func(r route) bool { return r.peer == peer })
+	if i < 0 {
+		return false
+	}
+
+	t.countKeptTime(o.routes[i].time, -1)
+	o.routes = slices.Delete(o.routes, i, i+1)
+
+	return true
+}
+
+// recordChange records that the pair of prefix and the AS of o changed, and
+// brings the count of origins without a kept route up to date.
+// wasWithoutKeptRoute says whether o was such an origin before the change. An
+// origin that no peer carries any more is removed.
+func (t *Table) recordChange(
+	prefix netip.Prefix, o *origin, wasWithoutKeptRoute bool,
+) {
+	t.changed[pair{prefix: prefix, asn: o.asn}] = struct{}{}
+
+	t.countOriginWithoutKeptRoute(o, wasWithoutKeptRoute)
+
+	if o.peers.None() {
+		t.removeOrigin(prefix, o)
+	}
+}
+
+// countOriginWithoutKeptRoute brings the count of origins without a kept route
+// up to date after a change to o. wasWithoutKeptRoute says whether o was such
+// an origin before the change.
+func (t *Table) countOriginWithoutKeptRoute(o *origin, wasWithoutKeptRoute bool) {
+	if wasWithoutKeptRoute {
+		t.originsWithoutKeptRoute--
+	}
+
+	if o.withoutKeptRoute() {
+		t.originsWithoutKeptRoute++
+	}
+}
+
+// countKeptTime adds delta to the number of kept routes with time at.
+func (t *Table) countKeptTime(at time.Time, delta int) {
+	count, _ := t.keptTimes.Get(keptTime{time: at})
+	count.time = at
+	count.routes += delta
+
+	if count.routes == 0 {
+		t.keptTimes.Delete(count)
+
+		return
+	}
+
+	t.keptTimes.ReplaceOrInsert(count)
+}
+
+// currentContent returns what the table holds for the pair now.
+func (t *Table) currentContent(key pair) Change {
+	o := t.byOrigin[key.asn][key.prefix]
+	if o == nil {
+		return Change{Prefix: key.prefix, OriginASN: key.asn, Removed: true}
+	}
+
+	return Change{
+		Prefix:    key.prefix,
+		OriginASN: key.asn,
+		Peers:     o.peers.Clone(),
+		Routes:    t.exportRoutes(o.routes),
+	}
+}
+
+// exportRoutes returns copies of routes that name each peer by its address, or
+// nil when there are none.
+func (t *Table) exportRoutes(routes []route) []Route {
+	if len(routes) == 0 {
+		return nil
+	}
+
+	exported := make([]Route, 0, len(routes))
+	for _, r := range routes {
+		exported = append(exported, Route{
+			Peer:   t.peerAddresses[r.peer],
+			ASPath: slices.Clone(r.path),
+			Time:   r.time,
+		})
+	}
+
+	return exported
+}
+
+// family is the part of the table for one IP version.
+type family struct {
+	// byLength[n] holds the prefixes of mask length n that some peer carries,
+	// each with its origins in AS number order.
+	byLength []map[netip.Prefix][]*origin
+	// minLookupLength is the shortest mask length the address lookup uses.
+	minLookupLength int
+	// routes counts the peer and prefix pairs: each peer that carries a
+	// prefix counts once for it.
+	routes int
+}
+
+// newFamily returns an empty family for addresses of bits bits.
+func newFamily(bits, minLookupLength int) family {
+	f := family{
+		byLength:        make([]map[netip.Prefix][]*origin, bits+1),
+		minLookupLength: minLookupLength,
+	}
+
+	for n := range f.byLength {
+		f.byLength[n] = make(map[netip.Prefix][]*origin)
+	}
+
+	return f
+}
+
+// prefixesByLength returns how many prefixes of each mask length f holds.
+func (f *family) prefixesByLength() []int {
+	counts := make([]int, len(f.byLength))
+	for n, prefixes := range f.byLength {
+		counts[n] = len(prefixes)
+	}
+
+	return counts
+}
+
+// origin is what the table holds for one prefix and origin AS.
+type origin struct {
+	asn int
+	// peers holds the number of each peer whose current route to the prefix
+	// has this origin AS.
+	peers bitset.BitSet
+	// routes are the newest of those routes, newest first, at most
+	// keptRoutesPerOrigin.
+	routes []route
+}
+
+// withoutKeptRoute says whether peers carry o but o keeps no route.
+func (o *origin) withoutKeptRoute() bool {
+	return o.peers.Any() && len(o.routes) == 0
+}
+
+// newestTime returns the time of the newest kept route of o, or nil when o
+// keeps none.
+func (o *origin) newestTime() *time.Time {
+	if len(o.routes) == 0 {
+		return nil
+	}
+
+	newest := o.routes[0].time
+
+	return &newest
+}
+
+// originWithPeer returns the origin whose peer set holds peer, or nil when
+// none does. A peer carries a prefix with one origin AS at most.
+func originWithPeer(origins []*origin, peer uint) *origin {
+	for _, o := range origins {
+		if o.peers.Test(peer) {
+			return o
+		}
+	}
+
+	return nil
+}
+
+// route is a kept route. peer is the peer's number.
+type route struct {
+	peer uint
+	path []int
+	time time.Time
+}
+
+// equal says whether r and other are the same peer's route with the same path
+// and time.
+func (r route) equal(other route) bool {
+	return r.peer == other.peer && r.time.Equal(other.time) &&
+		slices.Equal(r.path, other.path)
+}
+
+// pair names one prefix and origin AS.
+type pair struct {
+	prefix netip.Prefix
+	asn    int
+}
+
+// keptTime counts the kept routes with one time.
+type keptTime struct {
+	time   time.Time
+	routes int
+}

internal/routetable/routetable_test.go

@@ -0,0 +1,667 @@
+package routetable_test
+
+import (
+	"math"
+	"net/netip"
+	"reflect"
+	"slices"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/bits-and-blooms/bitset"
+
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+const (
+	peerA = "192.0.2.1"
+	peerB = "192.0.2.2"
+	peerC = "2001:db8::3"
+	peerD = "2001:db8::4"
+
+	asA = 64500
+	asB = 64501
+	asC = 64502
+	// transitAS is the first AS of every test path, the peer's own.
+	transitAS = 64496
+
+	prefixV4 = "10.1.2.0/24"
+	prefixV6 = "2001:db8:1::/48"
+)
+
+// at returns the time second seconds after the start of every test.
+func at(second int) time.Time {
+	return time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC).
+		Add(time.Duration(second) * time.Second)
+}
+
+// timeAt returns at(second) for a field that may be nil.
+func timeAt(second int) *time.Time {
+	t := at(second)
+
+	return &t
+}
+
+// routeOf returns the route peer announces with origin AS asn second seconds
+// after the start.
+func routeOf(peer string, asn, second int) routetable.Route {
+	return routetable.Route{
+		Peer: peer, ASPath: []int{transitAS, asn}, Time: at(second),
+	}
+}
+
+// step is one update a test applies to a table for one prefix.
+type step func(table *routetable.Table, prefix netip.Prefix)
+
+// announce is the step of peer announcing prefix with origin AS asn second
+// seconds after the start.
+func announce(peer string, asn, second int) step {
+	return func(table *routetable.Table, prefix netip.Prefix) {
+		table.Announce(prefix, asn, routeOf(peer, asn, second))
+	}
+}
+
+// withdraw is the step of peer withdrawing prefix.
+func withdraw(peer string) step {
+	return func(table *routetable.Table, prefix netip.Prefix) {
+		table.Withdraw(prefix, peer)
+	}
+}
+
+// peerDown is the step of peer's session going down.
+func peerDown(peer string) step {
+	return func(table *routetable.Table, _ netip.Prefix) {
+		table.PeerDown(peer)
+	}
+}
+
+// rule is one case of the rule that decides what the table holds for a prefix:
+// the steps applied to an empty table, and the prefix's origins after them.
+// The peers are numbered in the order they first announce, which is the order
+// Origin.Peers lists them in.
+type rule struct {
+	name  string
+	steps []step
+	want  []routetable.Origin
+}
+
+// checkRules runs each case for an IPv4 and an IPv6 prefix.
+func checkRules(t *testing.T, rules []rule) {
+	t.Helper()
+
+	for _, tc := range rules {
+		for _, text := range []string{prefixV4, prefixV6} {
+			t.Run(tc.name+" "+text, func(t *testing.T) {
+				t.Parallel()
+
+				prefix := netip.MustParsePrefix(text)
+				table := routetable.New(routetable.Params{})
+
+				for _, apply := range tc.steps {
+					apply(table, prefix)
+				}
+
+				got := table.Origins(prefix)
+				if !reflect.DeepEqual(got, tc.want) {
+					t.Errorf("Origins(%s) = %+v, want %+v", prefix, got, tc.want)
+				}
+			})
+		}
+	}
+}
+
+func TestAnnouncementsKeepTheTwoNewestRoutes(t *testing.T) {
+	t.Parallel()
+
+	checkRules(t, []rule{
+		{
+			name:  "an announcement is kept",
+			steps: []step{announce(peerA, asA, 10)},
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA},
+				Routes: []routetable.Route{routeOf(peerA, asA, 10)}}},
+		},
+		{
+			name: "a lagging peer's older route is not kept",
+			steps: []step{announce(peerA, asA, 20), announce(peerB, asA, 30),
+				announce(peerC, asA, 10)},
+			want: []routetable.Origin{{ASN: asA,
+				Peers: []string{peerA, peerB, peerC}, Routes: []routetable.Route{
+					routeOf(peerB, asA, 30), routeOf(peerA, asA, 20),
+				}}},
+		},
+		{
+			name: "a third newer route drops the older kept one, whose peer " +
+				"keeps its bit",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				announce(peerC, asA, 30)},
+			want: []routetable.Origin{{ASN: asA,
+				Peers: []string{peerA, peerB, peerC}, Routes: []routetable.Route{
+					routeOf(peerC, asA, 30), routeOf(peerB, asA, 20),
+				}}},
+		},
+		{
+			name: "a route as old as the older kept one replaces it",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				announce(peerC, asA, 10)},
+			want: []routetable.Origin{{ASN: asA,
+				Peers: []string{peerA, peerB, peerC}, Routes: []routetable.Route{
+					routeOf(peerB, asA, 20), routeOf(peerC, asA, 10),
+				}}},
+		},
+		{
+			name: "a peer's new route replaces its kept one",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				announce(peerA, asA, 30)},
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA, peerB},
+				Routes: []routetable.Route{
+					routeOf(peerA, asA, 30), routeOf(peerB, asA, 20),
+				}}},
+		},
+		{
+			name: "a peer's new route replaces its kept one even when older",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				announce(peerB, asA, 5)},
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA, peerB},
+				Routes: []routetable.Route{
+					routeOf(peerA, asA, 10), routeOf(peerB, asA, 5),
+				}}},
+		},
+	})
+}
+
+func TestAPeerCarriesAPrefixWithOneOrigin(t *testing.T) {
+	t.Parallel()
+
+	checkRules(t, []rule{
+		{
+			name: "a peer moving to a new origin leaves the old one",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				announce(peerA, asB, 30)},
+			want: []routetable.Origin{
+				{ASN: asA, Peers: []string{peerB},
+					Routes: []routetable.Route{routeOf(peerB, asA, 20)}},
+				{ASN: asB, Peers: []string{peerA},
+					Routes: []routetable.Route{routeOf(peerA, asB, 30)}},
+			},
+		},
+		{
+			name: "a peer whose route is not kept moves to a new origin",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				announce(peerC, asA, 30), announce(peerA, asB, 40)},
+			want: []routetable.Origin{
+				{ASN: asA, Peers: []string{peerB, peerC},
+					Routes: []routetable.Route{
+						routeOf(peerC, asA, 30), routeOf(peerB, asA, 20),
+					}},
+				{ASN: asB, Peers: []string{peerA},
+					Routes: []routetable.Route{routeOf(peerA, asB, 40)}},
+			},
+		},
+		{
+			name: "origins are listed in AS number order",
+			steps: []step{announce(peerA, asC, 10), announce(peerB, asA, 20),
+				announce(peerC, asB, 30)},
+			want: []routetable.Origin{
+				{ASN: asA, Peers: []string{peerB},
+					Routes: []routetable.Route{routeOf(peerB, asA, 20)}},
+				{ASN: asB, Peers: []string{peerC},
+					Routes: []routetable.Route{routeOf(peerC, asB, 30)}},
+				{ASN: asC, Peers: []string{peerA},
+					Routes: []routetable.Route{routeOf(peerA, asC, 10)}},
+			},
+		},
+	})
+}
+
+func TestWithdrawalsAndPeerDowns(t *testing.T) {
+	t.Parallel()
+
+	// threePeers is peerA, peerB and peerC announcing in turn; the routes of
+	// peerC and peerB are kept.
+	threePeers := []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+		announce(peerC, asA, 30)}
+
+	checkRules(t, []rule{
+		{
+			name:  "a withdrawal drops the peer's kept route",
+			steps: append(slices.Clone(threePeers), withdraw(peerC)),
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA, peerB},
+				Routes: []routetable.Route{routeOf(peerB, asA, 20)}}},
+		},
+		{
+			name:  "a withdrawal of a route that is not kept leaves the kept ones",
+			steps: append(slices.Clone(threePeers), withdraw(peerA)),
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerB, peerC},
+				Routes: []routetable.Route{
+					routeOf(peerC, asA, 30), routeOf(peerB, asA, 20),
+				}}},
+		},
+		{
+			name: "an origin whose kept routes are withdrawn keeps its other " +
+				"peers",
+			steps: append(slices.Clone(threePeers), withdraw(peerB),
+				withdraw(peerC)),
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA}}},
+		},
+		{
+			name: "an origin without kept routes keeps the next update",
+			steps: append(slices.Clone(threePeers), withdraw(peerB),
+				withdraw(peerC), announce(peerA, asA, 5)),
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA},
+				Routes: []routetable.Route{routeOf(peerA, asA, 5)}}},
+		},
+		{
+			name: "an origin no peer carries is removed",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asB, 20),
+				withdraw(peerA)},
+			want: []routetable.Origin{{ASN: asB, Peers: []string{peerB},
+				Routes: []routetable.Route{routeOf(peerB, asB, 20)}}},
+		},
+		{
+			name:  "a prefix no peer carries is removed",
+			steps: []step{announce(peerA, asA, 10), withdraw(peerA)},
+			want:  nil,
+		},
+		{
+			name:  "a withdrawal from a peer that does not carry it changes nothing",
+			steps: []step{announce(peerA, asA, 10), withdraw(peerB)},
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA},
+				Routes: []routetable.Route{routeOf(peerA, asA, 10)}}},
+		},
+		{
+			name: "a peer going down leaves the prefix",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				peerDown(peerB)},
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA},
+				Routes: []routetable.Route{routeOf(peerA, asA, 10)}}},
+		},
+		{
+			name: "a peer going down after it withdrew changes nothing",
+			steps: []step{announce(peerA, asA, 10), announce(peerB, asA, 20),
+				withdraw(peerB), peerDown(peerB)},
+			want: []routetable.Origin{{ASN: asA, Peers: []string{peerA},
+				Routes: []routetable.Route{routeOf(peerA, asA, 10)}}},
+		},
+	})
+}
+
+func TestPeerDownLeavesEveryPrefix(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+	prefixes := []netip.Prefix{
+		netip.MustParsePrefix(prefixV4),
+		netip.MustParsePrefix("10.2.0.0/16"),
+		netip.MustParsePrefix(prefixV6),
+	}
+
+	// peerA carries every prefix with origin asA, peerB the first one with
+	// origin asB.
+	for _, prefix := range prefixes {
+		table.Announce(prefix, asA, routeOf(peerA, asA, 10))
+	}
+
+	table.Announce(prefixes[0], asB, routeOf(peerB, asB, 20))
+	table.PeerDown(peerA)
+
+	want := []routetable.Origin{{ASN: asB, Peers: []string{peerB},
+		Routes: []routetable.Route{routeOf(peerB, asB, 20)}}}
+
+	got := table.Origins(prefixes[0])
+	if !reflect.DeepEqual(got, want) {
+		t.Errorf("Origins(%s) = %+v, want %+v", prefixes[0], got, want)
+	}
+
+	for _, prefix := range prefixes[1:] {
+		got := table.Origins(prefix)
+		if got != nil {
+			t.Errorf("Origins(%s) = %+v after the only peer went down, want nil",
+				prefix, got)
+		}
+	}
+
+	counts := table.Counts()
+	if counts.IPv4Prefixes != 1 || counts.IPv6Prefixes != 0 ||
+		counts.IPv4Routes != 1 || counts.IPv6Routes != 0 {
+		t.Errorf("Counts() = %+v, want one IPv4 prefix and route", counts)
+	}
+
+	if len(table.OriginatedPrefixes(asA)) != 0 {
+		t.Errorf("OriginatedPrefixes(%d) = %+v, want none",
+			asA, table.OriginatedPrefixes(asA))
+	}
+}
+
+func TestCountsOriginsWithoutKeptRoute(t *testing.T) {
+	t.Parallel()
+
+	prefix := netip.MustParsePrefix(prefixV4)
+	table := routetable.New(routetable.Params{})
+
+	steps := []struct {
+		apply step
+		want  int
+	}{
+		{announce(peerA, asA, 10), 0},
+		{announce(peerB, asA, 20), 0},
+		{announce(peerC, asA, 30), 0},
+		{withdraw(peerC), 0},
+		{withdraw(peerB), 1},
+		// An update from the last peer is kept.
+		{announce(peerA, asA, 40), 0},
+		{withdraw(peerA), 0},
+		// Once more without kept routes, then without peers: the origin
+		// is removed.
+		{announce(peerA, asA, 10), 0},
+		{announce(peerB, asA, 20), 0},
+		{announce(peerC, asA, 30), 0},
+		{withdraw(peerB), 0},
+		{withdraw(peerC), 1},
+		{peerDown(peerA), 0},
+	}
+
+	for i, s := range steps {
+		s.apply(table, prefix)
+
+		got := table.Counts().OriginsWithoutKeptRoute
+		if got != s.want {
+			t.Errorf("after step %d, OriginsWithoutKeptRoute = %d, want %d",
+				i, got, s.want)
+		}
+	}
+}
+
+// change is a routetable.Change with its peer set as a list of peer numbers.
+type change struct {
+	Prefix    string
+	OriginASN int
+	Removed   bool
+	Peers     []uint
+	Routes    []routetable.Route
+}
+
+// peerNumbers returns the numbers in set, or nil for a nil set.
+func peerNumbers(set *bitset.BitSet) []uint {
+	if set == nil {
+		return nil
+	}
+
+	return slices.Collect(set.EachSet())
+}
+
+func TestTakeChanges(t *testing.T) {
+	t.Parallel()
+
+	// peerA was numbered 5 before; the next new peer is 6.
+	table := routetable.New(routetable.Params{
+		PeerNumbers: map[string]uint{peerA: 5},
+	})
+	v4 := netip.MustParsePrefix(prefixV4)
+	v6 := netip.MustParsePrefix(prefixV6)
+
+	table.Announce(v6, asA, routeOf(peerA, asA, 10))
+	table.Announce(v4, asA, routeOf(peerA, asA, 10))
+	table.Announce(v4, asA, routeOf(peerB, asA, 20))
+
+	checkChanges(t, table, []change{
+		{Prefix: prefixV4, OriginASN: asA, Peers: []uint{5, 6},
+			Routes: []routetable.Route{
+				routeOf(peerB, asA, 20), routeOf(peerA, asA, 10),
+			}},
+		{Prefix: prefixV6, OriginASN: asA, Peers: []uint{5},
+			Routes: []routetable.Route{routeOf(peerA, asA, 10)}},
+	}, map[string]uint{peerB: 6})
+
+	// TakeChanges cleared the record.
+	checkChanges(t, table, []change{}, map[string]uint{})
+
+	// peerB moves to asB; peerA withdraws, which removes the pair of v4 and
+	// asA.
+	table.Announce(v4, asB, routeOf(peerB, asB, 30))
+	table.Withdraw(v4, peerA)
+
+	checkChanges(t, table, []change{
+		{Prefix: prefixV4, OriginASN: asA, Removed: true},
+		{Prefix: prefixV4, OriginASN: asB, Peers: []uint{6},
+			Routes: []routetable.Route{routeOf(peerB, asB, 30)}},
+	}, map[string]uint{})
+}
+
+func TestTakeChangesHoldsOnlyPairsThatChanged(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+	v4 := netip.MustParsePrefix(prefixV4)
+	v6 := netip.MustParsePrefix(prefixV6)
+
+	// peerA, peerB and peerC are numbered 0, 1 and 2. The routes of peerC and
+	// peerB to v4 are kept; peerA keeps only its bit.
+	announceAll(table, prefixV4, asA, 10, peerA, peerB, peerC)
+	table.Announce(v6, asA, routeOf(peerB, asA, 10))
+	table.TakeChanges(math.MaxInt)
+
+	// peerA announces again, with a time older than both kept routes.
+	table.Announce(v4, asA, routeOf(peerA, asA, 15))
+	checkChanges(t, table, []change{}, map[string]uint{})
+
+	// peerC announces its kept route again.
+	table.Announce(v4, asA, routeOf(peerC, asA, 30))
+	checkChanges(t, table, []change{}, map[string]uint{})
+
+	// peerC announces the same path with a newer time, so its kept route
+	// changes.
+	table.Announce(v4, asA, routeOf(peerC, asA, 40))
+	checkChanges(t, table, []change{
+		{Prefix: prefixV4, OriginASN: asA, Peers: []uint{0, 1, 2},
+			Routes: []routetable.Route{
+				routeOf(peerC, asA, 40), routeOf(peerB, asA, 20),
+			}},
+	}, map[string]uint{})
+
+	// peerA withdraws v6, which it does not carry.
+	table.Withdraw(v6, peerA)
+	checkChanges(t, table, []change{}, map[string]uint{})
+
+	// peerA going down changes v4, the one prefix it carries.
+	table.PeerDown(peerA)
+	checkChanges(t, table, []change{
+		{Prefix: prefixV4, OriginASN: asA, Peers: []uint{1, 2},
+			Routes: []routetable.Route{
+				routeOf(peerC, asA, 40), routeOf(peerB, asA, 20),
+			}},
+	}, map[string]uint{})
+
+	// peerA, which carries nothing now, goes down again.
+	table.PeerDown(peerA)
+	checkChanges(t, table, []change{}, map[string]uint{})
+}
+
+// TestTakeChangesTakesAtMostMaxChanges changes three prefix and origin AS
+// pairs, then takes the changes two at a time. The first call must return two
+// pairs and both new peers, and the second the third pair and no peer.
+func TestTakeChangesTakesAtMostMaxChanges(t *testing.T) {
+	t.Parallel()
+
+	const maxChanges = 2
+
+	table := routetable.New(routetable.Params{})
+	v4 := netip.MustParsePrefix(prefixV4)
+	v6 := netip.MustParsePrefix(prefixV6)
+
+	table.Announce(v4, asA, routeOf(peerA, asA, 10))
+	table.Announce(v4, asB, routeOf(peerB, asB, 10))
+	table.Announce(v6, asA, routeOf(peerA, asA, 10))
+
+	first, firstPeers := table.TakeChanges(maxChanges)
+	second, secondPeers := table.TakeChanges(maxChanges)
+
+	if len(first) != maxChanges || len(second) != 1 {
+		t.Fatalf("TakeChanges(%d) returned %d and then %d pairs, want %d and 1",
+			maxChanges, len(first), len(second), maxChanges)
+	}
+
+	wantPeers := map[string]uint{peerA: 0, peerB: 1}
+	if !reflect.DeepEqual(firstPeers, wantPeers) || len(secondPeers) != 0 {
+		t.Errorf("TakeChanges(%d) returned the new peers %v and then %v, "+
+			"want %v and none", maxChanges, firstPeers, secondPeers, wantPeers)
+	}
+
+	for _, c := range first {
+		if c.Prefix == second[0].Prefix && c.OriginASN == second[0].OriginASN {
+			t.Errorf("both calls returned %s with AS%d", c.Prefix, c.OriginASN)
+		}
+	}
+}
+
+// TestPutBackChanges takes a table's changes and puts them back, as a caller
+// does whose write failed. The next TakeChanges must return them again, with
+// each pair's content then, and with the changes since.
+func TestPutBackChanges(t *testing.T) {
+	t.Parallel()
+
+	table := routetable.New(routetable.Params{})
+	v4 := netip.MustParsePrefix(prefixV4)
+	v6 := netip.MustParsePrefix(prefixV6)
+
+	table.Announce(v4, asA, routeOf(peerA, asA, 10))
+	table.Announce(v6, asA, routeOf(peerA, asA, 10))
+	table.PutBackChanges(table.TakeChanges(math.MaxInt))
+
+	if got := table.ChangeCount(); got != 2 {
+		t.Errorf("ChangeCount() = %d after the changes were put back, want 2",
+			got)
+	}
+
+	// peerA withdraws v4, so v4's earlier content must not come back; peerB
+	// announces v6, and is a new peer as well.
+	table.Withdraw(v4, peerA)
+	table.Announce(v6, asA, routeOf(peerB, asA, 20))
+
+	checkChanges(t, table, []change{
+		{Prefix: prefixV4, OriginASN: asA, Removed: true},
+		{Prefix: prefixV6, OriginASN: asA, Peers: []uint{0, 1},
+			Routes: []routetable.Route{
+				routeOf(peerB, asA, 20), routeOf(peerA, asA, 10),
+			}},
+	}, map[string]uint{peerA: 0, peerB: 1})
+
+	if got := table.ChangeCount(); got != 0 {
+		t.Errorf("ChangeCount() = %d after TakeChanges, want 0", got)
+	}
+}
+
+func TestLoad(t *testing.T) {
+	t.Parallel()
+
+	const otherV4 = "10.9.0.0/16"
+
+	v4 := netip.MustParsePrefix(prefixV4)
+	v6 := netip.MustParsePrefix(prefixV6)
+	other := netip.MustParsePrefix(otherV4)
+
+	// The routes of peerC and peerB to v4 are kept; peerA keeps only its bit.
+	// The two routes to v6 have the same time. The routes of other were
+	// withdrawn while peerA still carries it.
+	source := routetable.New(routetable.Params{})
+	announceAll(source, prefixV4, asA, 10, peerA, peerB, peerC)
+	announceAll(source, prefixV4, asB, 40, peerD)
+	source.Announce(v6, asA, routeOf(peerA, asA, 50))
+	source.Announce(v6, asA, routeOf(peerC, asA, 50))
+	announceAll(source, otherV4, asB, 10, peerA, peerB, peerC)
+	source.Withdraw(other, peerB)
+	source.Withdraw(other, peerC)
+
+	changes, peers := source.TakeChanges(math.MaxInt)
+	loaded := routetable.New(routetable.Params{PeerNumbers: peers})
+
+	// Loading a pair twice changes nothing.
+	for range 2 {
+		for _, change := range changes {
+			loaded.Load(change)
+		}
+	}
+
+	for _, prefix := range []netip.Prefix{v4, v6, other} {
+		got, want := loaded.Origins(prefix), source.Origins(prefix)
+		if !reflect.DeepEqual(got, want) {
+			t.Errorf("Origins(%s) = %+v after the load, want %+v", prefix, got, want)
+		}
+	}
+
+	if !reflect.DeepEqual(loaded.Counts(), source.Counts()) {
+		t.Errorf("Counts() = %+v after the load, want %+v",
+			loaded.Counts(), source.Counts())
+	}
+
+	checkChanges(t, loaded, []change{}, map[string]uint{})
+}
+
+// TestConcurrentUse has goroutines update one table while others read it. The
+// race detector, which make test runs, fails it when a method uses the table
+// without its lock.
+func TestConcurrentUse(t *testing.T) {
+	t.Parallel()
+
+	const (
+		// rounds is how many times each goroutine calls its method.
+		rounds     = 1000
+		ipVersion  = 4
+		sampleSize = 10
+	)
+
+	table := routetable.New(routetable.Params{})
+	prefix := netip.MustParsePrefix(prefixV4)
+
+	uses := []func(round int){
+		func(round int) { table.Announce(prefix, asA, routeOf(peerA, asA, round)) },
+		func(round int) { table.Announce(prefix, asB, routeOf(peerB, asB, round)) },
+		func(int) { table.Withdraw(prefix, peerA) },
+		func(int) { table.PeerDown(peerB) },
+		func(int) { table.PutBackChanges(table.TakeChanges(math.MaxInt)) },
+		func(int) { table.ChangeCount() },
+		func(int) { table.Lookup(prefix.Addr()) },
+		func(int) { table.Origins(prefix) },
+		func(int) { table.OriginatedPrefixes(asA) },
+		func(int) { table.SamplePrefixes(ipVersion, prefix.Bits(), sampleSize) },
+		func(int) { table.Counts() },
+	}
+
+	var wg sync.WaitGroup
+	for _, use := range uses {
+		wg.Go(func() {
+			for round := range rounds {
+				use(round)
+			}
+		})
+	}
+
+	wg.Wait()
+}
+
+// checkChanges checks that TakeChanges, taking every change, returns want and
+// wantPeers.
+func checkChanges(
+	t *testing.T, table *routetable.Table, want []change,
+	wantPeers map[string]uint,
+) {
+	t.Helper()
+
+	changes, newPeers := table.TakeChanges(math.MaxInt)
+
+	got := make([]change, 0, len(changes))
+	for _, c := range changes {
+		got = append(got, change{
+			Prefix: c.Prefix.String(), OriginASN: c.OriginASN, Removed: c.Removed,
+			Peers: peerNumbers(c.Peers), Routes: c.Routes,
+		})
+	}
+
+	if !reflect.DeepEqual(got, want) {
+		t.Errorf("TakeChanges() changes = %+v, want %+v", got, want)
+	}
+
+	if !reflect.DeepEqual(newPeers, wantPeers) {
+		t.Errorf("TakeChanges() new peers = %v, want %v", newPeers, wantPeers)
+	}
+}

internal/routewatch/app.go

@@ -1,23 +1,34 @@
-// Package routewatch contains the primary RouteWatch type that represents a running instance
+// Package routewatch contains the primary RouteWatch type that represents a
-// of the application and contains pointers to its core dependencies, and is responsible for initialization.
+// running instance of the application and contains pointers to its core
+// dependencies, and is responsible for initialization. It also contains the
+// ASN WHOIS fetcher for background updates and the database maintainer for
+// background maintenance tasks.
 package routewatch
 
 import (
 	"context"
+	"errors"
 	"fmt"
+	"path/filepath"
 	"sync"
 	"time"
 
+	"git.eeqj.de/sneak/routewatch/internal/blocklist"
 	"git.eeqj.de/sneak/routewatch/internal/config"
 	"git.eeqj.de/sneak/routewatch/internal/database"
 	"git.eeqj.de/sneak/routewatch/internal/logger"
 	"git.eeqj.de/sneak/routewatch/internal/metrics"
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
 	"git.eeqj.de/sneak/routewatch/internal/server"
 	"git.eeqj.de/sneak/routewatch/internal/streamer"
 
 	"go.uber.org/fx"
 )
 
+// errNonBatchedNotImplemented is returned by Run when batched database writes
+// are disabled, since only the batched handlers exist.
+var errNonBatchedNotImplemented = errors.New("non-batched handlers not implemented")
+
 // Dependencies contains all dependencies for RouteWatch
 type Dependencies struct {
 	fx.In
@@ -45,8 +56,13 @@ type RouteWatch struct {
 	peeringHandler *PeeringHandler
 	asnFetcher     *ASNFetcher
 	dbMaintainer   *DBMaintainer
+	blocklists     *blocklist.Blocklists
 }
 
+// blocklistsDirName is the directory in the state directory that holds the
+// blocklist copies.
+const blocklistsDirName = "blocklists"
+
 // New creates a new RouteWatch instance
 func New(deps Dependencies) *RouteWatch {
 	rw := &RouteWatch{
@@ -68,14 +84,24 @@ func (rw *RouteWatch) Run(ctx context.Context) error {
 	// Apply runtime limit if specified
 	if rw.maxRuntime > 0 {
 		var cancel context.CancelFunc
+
 		ctx, cancel = context.WithTimeout(ctx, rw.maxRuntime)
 		defer cancel()
+
 		rw.logger.Info("Running with time limit", "max_runtime", rw.maxRuntime)
 	}
 
 	// Register database handler to process BGP UPDATE messages
 	if rw.config.EnableBatchedDatabaseWrites {
 		rw.logger.Info("Using batched database handlers for improved performance")
+
+		// The routing table is read before the streamer connects, so that
+		// the prefix handler applies the feed's updates to all of it.
+		table, err := rw.readRouteTable(ctx)
+		if err != nil {
+			return err
+		}
+
 		// ASHandler maintains the asns table
 		rw.dbHandler = NewASHandler(rw.db, rw.logger)
 		rw.streamer.RegisterHandler(rw.dbHandler)
@@ -84,8 +110,9 @@ func (rw *RouteWatch) Run(ctx context.Context) error {
 		rw.peerHandler = NewPeerHandler(rw.db, rw.logger)
 		rw.streamer.RegisterHandler(rw.peerHandler)
 
-		// PrefixHandler maintains the prefixes and live_routes tables
+		// PrefixHandler maintains the prefixes and live_routes tables, and
-		rw.prefixHandler = NewPrefixHandler(rw.db, rw.logger)
+		// the routing table in memory and its copy in the routes table
+		rw.prefixHandler = NewPrefixHandler(rw.db, rw.logger, table)
 		rw.prefixHandler.SetMetricsTracker(rw.streamer.GetMetricsTracker())
 		rw.streamer.RegisterHandler(rw.prefixHandler)
 
@@ -96,29 +123,44 @@ func (rw *RouteWatch) Run(ctx context.Context) error {
 		// Non-batched handlers not implemented yet
 		rw.logger.Error("Non-batched handlers not implemented")
 
-		return fmt.Errorf("non-batched handlers not implemented")
+		return errNonBatchedNotImplemented
 	}
 
 	// No longer need routing table handler - PrefixHandler maintains live_routes table
 
 	// Start streaming
-	if err := rw.streamer.Start(); err != nil {
+	err := rw.streamer.Start(ctx)
+	if err != nil {
 		return err
 	}
 
+	// Start ASN WHOIS fetcher for background updates. The HTTP handlers read
+	// the server's fetcher without a lock, so it is set before they run.
+	rw.asnFetcher = NewASNFetcher(rw.db, rw.logger.Logger)
+	rw.asnFetcher.Start(ctx)
+	rw.server.SetASNFetcher(rw.asnFetcher)
+
+	// Load the blocklist copies from disk and start downloading the lists.
+	// The HTTP handlers read the server's blocklists without a lock, so they
+	// are set before the handlers run.
+	blocklists := blocklist.New(blocklist.Params{
+		Dir:    filepath.Join(rw.config.StateDir, blocklistsDirName),
+		Lists:  blocklist.Lists(),
+		Logger: rw.logger.Logger,
+	})
+	blocklists.Start(ctx)
+	rw.blocklists = blocklists
+	rw.server.SetBlocklists(blocklists)
+
 	// Start HTTP server
-	if err := rw.server.Start(); err != nil {
+	err = rw.server.Start(ctx)
+	if err != nil {
 		return err
 	}
 
-	// Start ASN WHOIS fetcher for background updates
-	rw.asnFetcher = NewASNFetcher(rw.db, rw.logger.Logger)
-	rw.asnFetcher.Start()
-	rw.server.SetASNFetcher(rw.asnFetcher)
-
 	// Start database maintenance goroutine
 	rw.dbMaintainer = NewDBMaintainer(rw.db, rw.logger.Logger)
-	rw.dbMaintainer.Start()
+	rw.dbMaintainer.Start(ctx)
 
 	// Wait for context cancellation
 	<-ctx.Done()
@@ -126,56 +168,71 @@ func (rw *RouteWatch) Run(ctx context.Context) error {
 	return nil
 }
 
-// Shutdown performs graceful shutdown of all services
+// Shutdown stops all services in order: the streamer, the handlers, which
-func (rw *RouteWatch) Shutdown() {
+// write what they hold, the WHOIS fetcher, the database maintainer, the
+// blocklist downloads and the HTTP server, and last the database, which it
+// checkpoints and closes. Each stage waits at most until ctx ends, so
+// Shutdown returns soon after: a handler logs what it could not write by
+// then, and the HTTP server closes the connections of requests still
+// running.
+func (rw *RouteWatch) Shutdown(ctx context.Context) {
 	rw.mu.Lock()
 	if rw.shutdown {
 		rw.mu.Unlock()
 
 		return
 	}
+
 	rw.shutdown = true
 	rw.mu.Unlock()
 
-	// Stop batched handlers first to flush remaining batches
+	// Stop the streamer first: it returns once the handlers have been handed
+	// every message it queued, or when ctx ends. Then each handler writes
+	// what it holds.
+	rw.streamer.Stop(ctx)
+
 	if rw.dbHandler != nil {
 		rw.logger.Info("Flushing database handler")
-		rw.dbHandler.Stop()
+		rw.dbHandler.Stop(ctx)
 	}
+
 	if rw.peerHandler != nil {
 		rw.logger.Info("Flushing peer handler")
-		rw.peerHandler.Stop()
+		rw.peerHandler.Stop(ctx)
 	}
+
 	if rw.prefixHandler != nil {
 		rw.logger.Info("Flushing prefix handler")
-		rw.prefixHandler.Stop()
+		rw.prefixHandler.Stop(ctx)
 	}
+
 	if rw.peeringHandler != nil {
 		rw.logger.Info("Flushing peering handler")
-		rw.peeringHandler.Stop()
+		rw.peeringHandler.Stop(ctx)
 	}
 
 	// Stop ASN WHOIS fetcher
 	if rw.asnFetcher != nil {
-		rw.asnFetcher.Stop()
+		rw.asnFetcher.Stop(ctx)
 	}
 
 	// Stop database maintainer
 	if rw.dbMaintainer != nil {
-		rw.dbMaintainer.Stop()
+		rw.dbMaintainer.Stop(ctx)
 	}
 
-	// Stop services
+	if rw.blocklists != nil {
-	rw.streamer.Stop()
+		rw.blocklists.Stop(ctx)
-
-	// Stop HTTP server with a timeout
-	const serverStopTimeout = 5 * time.Second
-	stopCtx, cancel := context.WithTimeout(context.Background(), serverStopTimeout)
-	defer cancel()
-	if err := rw.server.Stop(stopCtx); err != nil {
-		rw.logger.Error("Failed to stop HTTP server gracefully", "error", err)
 	}
 
+	err := rw.server.Stop(ctx)
+	if err != nil {
+		rw.logger.Error("Stopped the HTTP server before every request finished",
+			"error", err)
+	}
+
+	rw.closeDatabase(ctx)
+
 	// Log final metrics
 	metrics := rw.streamer.GetMetrics()
 	rw.logger.Info("Final metrics",
@@ -185,12 +242,53 @@ func (rw *RouteWatch) Shutdown() {
 		"bits_per_sec", metrics.BitsPerSec,
 		"duration", time.Since(metrics.ConnectedSince),
 	)
+}
 
+// closeDatabase checkpoints and closes the database, unless ctx has ended: a
+// write may then still be running, and both would wait for it. SQLite rolls
+// back a write cut off by the exit when the database next opens.
+func (rw *RouteWatch) closeDatabase(ctx context.Context) {
+	if ctx.Err() != nil {
+		rw.logger.Error("Stop deadline passed; the database is not closed")
+
+		return
+	}
+
+	err := rw.db.Checkpoint(ctx)
+	if err != nil {
+		rw.logger.Error("Failed to checkpoint the database", "error", err)
+	}
+
+	err = rw.db.Close()
+	if err != nil {
+		rw.logger.Error("Failed to close the database", "error", err)
+	}
+}
+
+// readRouteTable reads the routing table from the database and logs how many
+// prefix and origin AS pairs and peers it holds and how long the read took.
+// ctx is fx's start context, which ends 15 seconds after the start. Reading a
+// full table can take longer, so the read does not stop with ctx.
+func (rw *RouteWatch) readRouteTable(
+	ctx context.Context,
+) (*routetable.Table, error) {
+	start := time.Now()
+
+	table, err := rw.db.ReadRouteTable(context.WithoutCancel(ctx))
+	if err != nil {
+		return nil, fmt.Errorf("failed to read the routing table: %w", err)
+	}
+
+	counts := table.Counts()
+	rw.logger.Info("Read the routing table", "pairs", counts.Pairs,
+		"peers", counts.Peers, "duration_ms", time.Since(start).Milliseconds())
+
+	return table, nil
 }
 
 // getModule provides all fx dependencies
-func getModule() fx.Option {
+func getModule() []fx.Option {
-	return fx.Options(
+	return []fx.Option{
 		fx.Provide(
 			logger.New,
 			config.New,
@@ -203,5 +301,5 @@ func getModule() fx.Option {
 			server.New,
 			New,
 		),
-	)
+	}
 }

internal/routewatch/app_integration_internal_test.go

@@ -0,0 +1,866 @@
+package routewatch
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net"
+	"net/http"
+	"net/http/httptest"
+	"slices"
+	"strconv"
+	"strings"
+	"sync"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/config"
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/metrics"
+	"git.eeqj.de/sneak/routewatch/internal/ristypes"
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+	"git.eeqj.de/sneak/routewatch/internal/server"
+	"git.eeqj.de/sneak/routewatch/internal/streamer"
+)
+
+// mockStore is a mock implementation of database.Store for testing
+type mockStore struct {
+	mu sync.Mutex
+
+	// Counters for tracking calls
+	ASNCount        int
+	PrefixCount     int
+	PeeringCount    int
+	RouteCount      int
+	WithdrawalCount int
+
+	// Track unique items
+	ASNs     map[int]*database.ASN
+	Prefixes map[string]*database.Prefix
+	Peerings map[string]bool // key is "from_to"
+	Routes   map[string]bool // key is "prefix_origin_peer"
+	// TablePrefixes holds the prefixes that the written changes of the
+	// routing table leave some peer carrying.
+	TablePrefixes map[string]bool
+
+	// Track IP versions
+	IPv4Prefixes int
+	IPv6Prefixes int
+}
+
+// newMockStore creates a new mock store
+func newMockStore() *mockStore {
+	return &mockStore{
+		ASNs:          make(map[int]*database.ASN),
+		Prefixes:      make(map[string]*database.Prefix),
+		Peerings:      make(map[string]bool),
+		Routes:        make(map[string]bool),
+		TablePrefixes: make(map[string]bool),
+	}
+}
+
+// GetOrCreateASN mock implementation
+func (m *mockStore) GetOrCreateASN(
+	number int, timestamp time.Time,
+) (*database.ASN, error) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	if asn, exists := m.ASNs[number]; exists {
+		asn.LastSeen = timestamp
+
+		return asn, nil
+	}
+
+	asn := &database.ASN{
+		ASN:       number,
+		FirstSeen: timestamp,
+		LastSeen:  timestamp,
+	}
+	m.ASNs[number] = asn
+	m.ASNCount++
+
+	return asn, nil
+}
+
+// UpdatePrefixesBatch mock implementation
+func (m *mockStore) UpdatePrefixesBatch(prefixes map[string]time.Time) error {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	for prefix, timestamp := range prefixes {
+		if _, exists := m.Prefixes[prefix]; !exists {
+			const (
+				ipVersionV4 = 4
+				ipVersionV6 = 6
+			)
+
+			ipVersion := ipVersionV4
+			if strings.Contains(prefix, ":") {
+				ipVersion = ipVersionV6
+			}
+
+			m.Prefixes[prefix] = &database.Prefix{
+				Prefix:    prefix,
+				IPVersion: ipVersion,
+				FirstSeen: timestamp,
+			}
+		}
+	}
+
+	return nil
+}
+
+// GetOrCreatePrefix mock implementation
+func (m *mockStore) GetOrCreatePrefix(
+	prefix string, timestamp time.Time,
+) (*database.Prefix, error) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	if p, exists := m.Prefixes[prefix]; exists {
+		return p, nil
+	}
+
+	const (
+		ipVersionV4 = 4
+		ipVersionV6 = 6
+	)
+
+	ipVersion := ipVersionV4
+	if strings.Contains(prefix, ":") {
+		ipVersion = ipVersionV6
+	}
+
+	p := &database.Prefix{
+		Prefix:    prefix,
+		IPVersion: ipVersion,
+		FirstSeen: timestamp,
+	}
+	m.Prefixes[prefix] = p
+	m.PrefixCount++
+
+	if ipVersion == ipVersionV4 {
+		m.IPv4Prefixes++
+	} else {
+		m.IPv6Prefixes++
+	}
+
+	return p, nil
+}
+
+// RecordAnnouncement mock implementation
+func (m *mockStore) RecordAnnouncement(_ *database.Announcement) error {
+	// Not tracking announcements in detail for now
+	return nil
+}
+
+// RecordPeering mock implementation
+func (m *mockStore) RecordPeering(asA, asB int, _ database.SeenTimes) error {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	// Normalize
+	if asA > asB {
+		asA, asB = asB, asA
+	}
+
+	key := fmt.Sprintf("%d_%d", asA, asB)
+	if !m.Peerings[key] {
+		m.Peerings[key] = true
+		m.PeeringCount++
+	}
+
+	return nil
+}
+
+// UpdatePeer mock implementation
+func (m *mockStore) UpdatePeer(_ string, _ int, _ string, _ time.Time) error {
+	// Simple mock - just return nil
+	return nil
+}
+
+// Close mock implementation
+func (m *mockStore) Close() error {
+	return nil
+}
+
+// GetStats returns statistics about the mock store
+func (m *mockStore) GetStats() (database.Stats, error) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	return database.Stats{
+		ASNs:         len(m.ASNs),
+		Prefixes:     len(m.Prefixes),
+		IPv4Prefixes: m.IPv4Prefixes,
+		IPv6Prefixes: m.IPv6Prefixes,
+		Peerings:     m.PeeringCount,
+		Peers:        10, // Mock peer count
+	}, nil
+}
+
+// GetStatsContext returns statistics about the mock store with context support
+func (m *mockStore) GetStatsContext(_ context.Context) (database.Stats, error) {
+	return m.GetStats()
+}
+
+// UpsertLiveRoute mock implementation
+func (m *mockStore) UpsertLiveRoute(_ *database.LiveRoute) error {
+	// Simple mock - just return nil
+	return nil
+}
+
+// DeleteLiveRoute mock implementation
+func (m *mockStore) DeleteLiveRoute(_, _ string) error {
+	// Simple mock - just return nil
+	return nil
+}
+
+// GetPrefixDistribution mock implementation
+func (m *mockStore) GetPrefixDistribution() (
+	[]database.PrefixDistribution, []database.PrefixDistribution, error,
+) {
+	// Return empty distributions for now
+	return nil, nil, nil
+}
+
+// GetPrefixDistributionContext mock implementation with context support
+func (m *mockStore) GetPrefixDistributionContext(_ context.Context) (
+	[]database.PrefixDistribution, []database.PrefixDistribution, error,
+) {
+	return m.GetPrefixDistribution()
+}
+
+// GetLiveRouteCounts mock implementation
+func (m *mockStore) GetLiveRouteCounts() (int, int, error) {
+	// Return mock counts
+	return m.RouteCount / 2, m.RouteCount / 2, nil
+}
+
+// GetLiveRouteCountsContext mock implementation with context support
+func (m *mockStore) GetLiveRouteCountsContext(_ context.Context) (int, int, error) {
+	return m.GetLiveRouteCounts()
+}
+
+// GetASDetails mock implementation
+func (m *mockStore) GetASDetails(asn int) (*database.ASN, []database.LiveRoute, error) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	// Check if ASN exists
+	if asnInfo, exists := m.ASNs[asn]; exists {
+		// Return empty prefixes for now
+		return asnInfo, []database.LiveRoute{}, nil
+	}
+
+	return nil, nil, database.ErrNoRoute
+}
+
+// GetASDetailsContext mock implementation with context support
+func (m *mockStore) GetASDetailsContext(
+	_ context.Context, asn int,
+) (*database.ASN, []database.LiveRoute, error) {
+	return m.GetASDetails(asn)
+}
+
+// GetPrefixDetails mock implementation
+func (m *mockStore) GetPrefixDetails(
+	_ string,
+) ([]database.LiveRoute, map[int]database.ASN, error) {
+	// Return empty routes for now
+	return []database.LiveRoute{}, map[int]database.ASN{}, nil
+}
+
+// GetPrefixDetailsContext mock implementation with context support
+func (m *mockStore) GetPrefixDetailsContext(
+	_ context.Context, prefix string,
+) ([]database.LiveRoute, map[int]database.ASN, error) {
+	return m.GetPrefixDetails(prefix)
+}
+
+func (m *mockStore) GetRandomPrefixesByLength(
+	_, _, _ int,
+) ([]database.LiveRoute, error) {
+	// Return empty routes for now
+	return []database.LiveRoute{}, nil
+}
+
+// GetRandomPrefixesByLengthContext mock implementation with context support
+func (m *mockStore) GetRandomPrefixesByLengthContext(
+	_ context.Context, maskLength, ipVersion, limit int,
+) ([]database.LiveRoute, error) {
+	return m.GetRandomPrefixesByLength(maskLength, ipVersion, limit)
+}
+
+// GetASPeers mock implementation
+func (m *mockStore) GetASPeers(_ int) ([]database.ASPeer, error) {
+	// Return empty peers for now
+	return []database.ASPeer{}, nil
+}
+
+// GetASPeersContext mock implementation with context support
+func (m *mockStore) GetASPeersContext(
+	_ context.Context, asn int,
+) ([]database.ASPeer, error) {
+	return m.GetASPeers(asn)
+}
+
+// GetIPInfo mock implementation
+func (m *mockStore) GetIPInfo(ip string) (*database.IPInfo, error) {
+	return m.GetIPInfoContext(context.Background(), ip)
+}
+
+// GetIPInfoContext mock implementation with context support
+func (m *mockStore) GetIPInfoContext(
+	_ context.Context, ip string,
+) (*database.IPInfo, error) {
+	now := time.Now()
+
+	return &database.IPInfo{
+		IP:          ip,
+		Netblock:    "8.8.8.0/24",
+		MaskLength:  24,
+		IPVersion:   4,
+		NumPeers:    3,
+		ASN:         15169,
+		Handle:      "GOOGLE",
+		Description: "Google LLC",
+		CountryCode: "US",
+		FirstSeen:   now.Add(-24 * time.Hour),
+		LastSeen:    now,
+	}, nil
+}
+
+// GetNextStaleASN mock implementation
+func (m *mockStore) GetNextStaleASN(_ context.Context, _ time.Duration) (int, error) {
+	return 0, database.ErrNoStaleASN
+}
+
+// UpdateASNWHOIS mock implementation
+func (m *mockStore) UpdateASNWHOIS(
+	_ context.Context, _ *database.ASNWHOISUpdate,
+) error {
+	return nil
+}
+
+// GetWHOISStats mock implementation
+func (m *mockStore) GetWHOISStats(
+	_ context.Context, _ time.Duration,
+) (*database.WHOISStats, error) {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	return &database.WHOISStats{
+		TotalASNs:    len(m.ASNs),
+		FreshASNs:    0,
+		StaleASNs:    0,
+		NeverFetched: len(m.ASNs),
+	}, nil
+}
+
+// UpsertLiveRouteBatch mock implementation
+func (m *mockStore) UpsertLiveRouteBatch(routes []*database.LiveRoute) error {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	for _, route := range routes {
+		// Track prefix
+		if _, exists := m.Prefixes[route.Prefix]; !exists {
+			m.Prefixes[route.Prefix] = &database.Prefix{
+				Prefix:    route.Prefix,
+				IPVersion: route.IPVersion,
+				FirstSeen: route.LastUpdated,
+			}
+			m.PrefixCount++
+
+			if route.IPVersion == 4 {
+				m.IPv4Prefixes++
+			} else {
+				m.IPv6Prefixes++
+			}
+		}
+
+		m.RouteCount++
+	}
+
+	return nil
+}
+
+// DeleteLiveRouteBatch mock implementation
+func (m *mockStore) DeleteLiveRouteBatch(_ []database.LiveRouteDeletion) error {
+	// Simple mock - just return nil
+	return nil
+}
+
+// DeleteLiveRoutesOfPeer mock implementation
+func (m *mockStore) DeleteLiveRoutesOfPeer(_ string) (int, error) {
+	// Simple mock - deletes nothing
+	return 0, nil
+}
+
+// WriteRouteChanges records the prefixes the changes leave some peer carrying,
+// whatever their origin AS.
+func (m *mockStore) WriteRouteChanges(
+	changes []routetable.Change, _ map[string]uint,
+) error {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	for _, change := range changes {
+		if change.Removed {
+			delete(m.TablePrefixes, change.Prefix.String())
+		} else {
+			m.TablePrefixes[change.Prefix.String()] = true
+		}
+	}
+
+	return nil
+}
+
+// ReadRouteTable returns an empty routing table.
+func (m *mockStore) ReadRouteTable(_ context.Context) (*routetable.Table, error) {
+	return routetable.New(routetable.Params{}), nil
+}
+
+// GetOrCreateASNBatch mock implementation
+func (m *mockStore) GetOrCreateASNBatch(asns map[int]database.SeenTimes) error {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+
+	for number, seen := range asns {
+		if _, exists := m.ASNs[number]; !exists {
+			m.ASNs[number] = &database.ASN{
+				ASN:       number,
+				FirstSeen: seen.First,
+				LastSeen:  seen.Last,
+			}
+			m.ASNCount++
+		}
+	}
+
+	return nil
+}
+
+// UpdatePeerBatch mock implementation
+func (m *mockStore) UpdatePeerBatch(_ map[string]database.PeerUpdate) error {
+	// Simple mock - just return nil
+	return nil
+}
+
+// Vacuum mock implementation
+func (m *mockStore) Vacuum(_ context.Context) error {
+	return nil
+}
+
+// Analyze mock implementation
+func (m *mockStore) Analyze(_ context.Context) error {
+	return nil
+}
+
+// Checkpoint mock implementation
+func (m *mockStore) Checkpoint(_ context.Context) error {
+	return nil
+}
+
+// Ping mock implementation
+func (m *mockStore) Ping(_ context.Context) error {
+	return nil
+}
+
+func TestRouteWatchLiveFeed(t *testing.T) {
+	t.Parallel()
+
+	if testing.Short() {
+		t.Skip("skipping live RIPE RIS network feed test in short mode; " +
+			"run without -short to include it")
+	}
+
+	// Create mock database
+	mockDB := newMockStore()
+
+	defer func() {
+		err := mockDB.Close()
+		if err != nil {
+			t.Errorf("failed to close mock database: %v", err)
+		}
+	}()
+
+	logger := logger.New()
+
+	// Create metrics tracker
+	metricsTracker := metrics.New()
+
+	// Create streamer
+	s := streamer.New(logger, metricsTracker)
+
+	// The state directory holds the blocklist copies Run downloads.
+	cfg := &config.Config{
+		StateDir:                    t.TempDir(),
+		MaxRuntime:                  5 * time.Second,
+		EnableBatchedDatabaseWrites: true,
+	}
+
+	// Create server
+	srv := server.New(mockDB, s, logger, cfg)
+
+	// Create RouteWatch with 5 second limit
+	deps := Dependencies{
+		DB:       mockDB,
+		Streamer: s,
+		Server:   srv,
+		Logger:   logger,
+		Config:   cfg,
+	}
+	rw := New(deps)
+
+	// Run with context
+	ctx := context.Background()
+	go func() {
+		_ = rw.Run(ctx)
+	}()
+
+	// Wait for the configured duration
+	time.Sleep(5 * time.Second)
+
+	// Force peering processing for test
+	if rw.peeringHandler != nil {
+		rw.peeringHandler.ProcessPeeringsNow()
+	}
+
+	// Get statistics
+	stats, err := mockDB.GetStats()
+	if err != nil {
+		t.Fatalf("Failed to get stats: %v", err)
+	}
+
+	if stats.ASNs == 0 {
+		t.Error("Expected to receive some ASNs from live feed")
+	}
+
+	t.Logf("Received %d unique ASNs in 5 seconds", stats.ASNs)
+
+	if stats.Prefixes == 0 {
+		t.Error("Expected to receive some prefixes from live feed")
+	}
+
+	t.Logf("Received %d unique prefixes (%d IPv4, %d IPv6) in 5 seconds",
+		stats.Prefixes, stats.IPv4Prefixes, stats.IPv6Prefixes)
+
+	if stats.Peerings == 0 {
+		t.Error("Expected to receive some peerings from live feed")
+	}
+
+	t.Logf("Recorded %d AS peering relationships in 5 seconds", stats.Peerings)
+}
+
+// gatedPrefixHandler is a PrefixHandler whose worker in the streamer waits on
+// the first message until release is closed, so the messages after it stay on
+// the streamer's queue.
+type gatedPrefixHandler struct {
+	*PrefixHandler
+
+	release chan struct{}
+}
+
+func (h gatedPrefixHandler) HandleMessage(msg *ristypes.RISMessage) {
+	<-h.release
+	h.PrefixHandler.HandleMessage(msg)
+}
+
+// announcementFeed returns the lines RIS Live would send for one UPDATE per
+// prefix, each announcing that prefix from testPeerIP with origin testASNA,
+// followed by a KEEPALIVE. The streamer hands a KEEPALIVE to no handler, so
+// once a raw handler has seen it, every update before it is queued.
+func announcementFeed(prefixes []string) string {
+	var feed strings.Builder
+
+	for _, prefix := range prefixes {
+		fmt.Fprintf(&feed, `{"type":"ris_message","data":{"type":"UPDATE",`+
+			`"peer":"%s","peer_asn":"%d","timestamp":1700000000,"path":[%d],`+
+			`"announcements":[{"next_hop":"%s","prefixes":["%s"]}]}}`+"\n",
+			testPeerIP, testASNA, testASNA, testPeerIP, prefix)
+	}
+
+	fmt.Fprintf(&feed, `{"type":"ris_message","data":{"type":"KEEPALIVE",`+
+		`"peer":"%s","peer_asn":"%d","timestamp":1700000000}}`+"\n",
+		testPeerIP, testASNA)
+
+	return feed.String()
+}
+
+// TestShutdownWritesTheUpdatesQueuedInTheStreamer has a local server send
+// announcements of three prefixes and keep the response open. The prefix
+// handler's worker waits on the first one until the server sees the streamer
+// close the connection, which only Shutdown does, so all three are still in
+// the streamer when Shutdown starts. Each prefix must be in the database when
+// Shutdown returns, both in the prefix tables and in the routing table's
+// writes.
+func TestShutdownWritesTheUpdatesQueuedInTheStreamer(t *testing.T) {
+	t.Parallel()
+
+	const maxWait = 10 * time.Second
+
+	prefixes := []string{"192.0.2.0/24", "198.51.100.0/24", "203.0.113.0/24"}
+	feed := announcementFeed(prefixes)
+
+	release := make(chan struct{})
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, r *http.Request) {
+			_, _ = io.WriteString(w, feed)
+			_ = http.NewResponseController(w).Flush()
+
+			<-r.Context().Done()
+			close(release)
+		}))
+	defer srv.Close()
+
+	store := newMockStore()
+	log := logger.New()
+	cfg := &config.Config{}
+	s := streamer.New(log, metrics.New())
+	s.SetURL(srv.URL)
+
+	rw := New(Dependencies{
+		DB:       store,
+		Streamer: s,
+		Server:   server.New(store, s, log, cfg),
+		Logger:   log,
+		Config:   cfg,
+	})
+	rw.prefixHandler = NewPrefixHandler(store, log, routetable.New(routetable.Params{}))
+	s.RegisterHandler(gatedPrefixHandler{
+		PrefixHandler: rw.prefixHandler, release: release,
+	})
+
+	linesSeen := make(chan struct{}, len(prefixes)+1)
+
+	s.RegisterRawHandler(func(string) { linesSeen <- struct{}{} })
+
+	err := s.Start(t.Context())
+	if err != nil {
+		t.Fatalf("Start returned error: %v", err)
+	}
+
+	for range cap(linesSeen) {
+		select {
+		case <-linesSeen:
+		case <-time.After(maxWait):
+			t.Fatalf("the streamer did not read %d lines within %s",
+				cap(linesSeen), maxWait)
+		}
+	}
+
+	shutDown := make(chan struct{})
+
+	go func() {
+		rw.Shutdown(t.Context())
+		close(shutDown)
+	}()
+
+	select {
+	case <-shutDown:
+	case <-time.After(maxWait):
+		t.Fatalf("Shutdown did not return within %s", maxWait)
+	}
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	for _, prefix := range prefixes {
+		_, inPrefixTables := store.Prefixes[prefix]
+		if !inPrefixTables || !store.TablePrefixes[prefix] {
+			t.Errorf("%s was queued in the streamer when Shutdown started; "+
+				"written to the prefix tables %t, in the routing table's "+
+				"writes %t", prefix, inPrefixTables, store.TablePrefixes[prefix])
+		}
+	}
+}
+
+// freePort returns a TCP port that was free a moment ago, for a server the
+// test starts.
+func freePort(t *testing.T) int {
+	t.Helper()
+
+	var listenConfig net.ListenConfig
+
+	listener, err := listenConfig.Listen(t.Context(), "tcp", ":0")
+	if err != nil {
+		t.Fatalf("failed to find a free port: %v", err)
+	}
+	defer func() { _ = listener.Close() }()
+
+	addr, ok := listener.Addr().(*net.TCPAddr)
+	if !ok {
+		t.Fatalf("listener address %v is not a TCP address", listener.Addr())
+	}
+
+	return addr.Port
+}
+
+// TestShutdownReturnsByItsDeadlineWhileAWriteIsBlocked starts the HTTP server
+// and has the AS handler start a write that the store holds until the test
+// ends, past Shutdown's deadline. Shutdown must return within its deadline
+// plus a small margin, with the HTTP server stopped, and must neither
+// checkpoint nor close the database: the real one would wait for the write.
+// Before the fix Shutdown waited for the write.
+func TestShutdownReturnsByItsDeadlineWhileAWriteIsBlocked(t *testing.T) {
+	t.Parallel()
+
+	const (
+		deadline = 100 * time.Millisecond
+		// margin is how long Shutdown may take after its deadline, to stop
+		// the HTTP server.
+		margin = 2 * time.Second
+	)
+
+	blocking := newBlockingStore()
+	defer blocking.releaseWrites()
+
+	store := &closeOrderStore{Store: blocking}
+	log := logger.New()
+	port := freePort(t)
+	cfg := &config.Config{Port: port}
+	s := streamer.New(log, metrics.New())
+
+	rw := New(Dependencies{
+		DB:       store,
+		Streamer: s,
+		Server:   server.New(store, s, log, cfg),
+		Logger:   log,
+		Config:   cfg,
+	})
+
+	err := rw.server.Start(t.Context())
+	if err != nil {
+		t.Fatalf("failed to start the HTTP server: %v", err)
+	}
+
+	rw.dbHandler = NewASHandler(store, log)
+	fill := ristypes.ASPath{Sequence: slices.Repeat([]int{testASNA}, asnBatchSize)}
+	startBlockedWrite(t, blocking, rw.dbHandler,
+		[]*ristypes.RISMessage{{Path: fill}})
+
+	ctx, cancel := context.WithTimeout(t.Context(), deadline)
+	defer cancel()
+
+	shutDown := make(chan struct{})
+
+	go func() {
+		rw.Shutdown(ctx)
+		close(shutDown)
+	}()
+
+	select {
+	case <-shutDown:
+	case <-time.After(deadline + margin):
+		t.Fatalf("Shutdown did not return within %s of its deadline while "+
+			"a write was blocked", margin)
+	}
+
+	var dialer net.Dialer
+
+	addr := net.JoinHostPort("127.0.0.1", strconv.Itoa(port))
+
+	conn, err := dialer.DialContext(t.Context(), "tcp", addr)
+	if err == nil {
+		_ = conn.Close()
+
+		t.Errorf("the HTTP server still accepts connections on %s after "+
+			"Shutdown returned", addr)
+	}
+
+	store.callsMu.Lock()
+	defer store.callsMu.Unlock()
+
+	if slices.Contains(store.calls, callCheckpoint) ||
+		slices.Contains(store.calls, callClose) {
+		t.Errorf("Shutdown made the calls %q past its deadline, want no %s "+
+			"and no %s", store.calls, callCheckpoint, callClose)
+	}
+}
+
+// closeOrderStore is a store that records, in order, the AS writes,
+// checkpoints and closes it is given. It passes the AS writes on to Store.
+type closeOrderStore struct {
+	database.Store
+
+	callsMu sync.Mutex
+	calls   []string
+}
+
+// The calls closeOrderStore records.
+const (
+	callWrite      = "AS write"
+	callCheckpoint = "checkpoint"
+	callClose      = "close"
+)
+
+func (s *closeOrderStore) GetOrCreateASNBatch(
+	asns map[int]database.SeenTimes,
+) error {
+	s.record(callWrite)
+
+	return s.Store.GetOrCreateASNBatch(asns)
+}
+
+func (s *closeOrderStore) Checkpoint(_ context.Context) error {
+	s.record(callCheckpoint)
+
+	return nil
+}
+
+func (s *closeOrderStore) Close() error {
+	s.record(callClose)
+
+	return nil
+}
+
+func (s *closeOrderStore) record(call string) {
+	s.callsMu.Lock()
+	defer s.callsMu.Unlock()
+
+	s.calls = append(s.calls, call)
+}
+
+// TestShutdownCheckpointsAndClosesTheDatabaseLast hands the AS handler one
+// message and shuts down with time to spare. The handler's last write must
+// come first, then a checkpoint, then the close. Before the fix Shutdown
+// neither checkpointed nor closed the database.
+func TestShutdownCheckpointsAndClosesTheDatabaseLast(t *testing.T) {
+	t.Parallel()
+
+	const deadline = 10 * time.Second
+
+	store := &closeOrderStore{Store: newMockStore()}
+	log := logger.New()
+	cfg := &config.Config{}
+	s := streamer.New(log, metrics.New())
+
+	rw := New(Dependencies{
+		DB:       store,
+		Streamer: s,
+		Server:   server.New(store, s, log, cfg),
+		Logger:   log,
+		Config:   cfg,
+	})
+	rw.dbHandler = NewASHandler(store, log)
+	rw.dbHandler.HandleMessage(&ristypes.RISMessage{
+		Path: ristypes.ASPath{Sequence: []int{testASNA}},
+	})
+
+	ctx, cancel := context.WithTimeout(t.Context(), deadline)
+	defer cancel()
+
+	rw.Shutdown(ctx)
+
+	store.callsMu.Lock()
+	defer store.callsMu.Unlock()
+
+	want := []string{callWrite, callCheckpoint, callClose}
+	if !slices.Equal(store.calls, want) {
+		t.Errorf("Shutdown made the calls %q, want %q", store.calls, want)
+	}
+}

internal/routewatch/app_integration_test.go

@@ -1,481 +0,0 @@
-package routewatch
-
-import (
-	"context"
-	"fmt"
-	"strings"
-	"sync"
-	"testing"
-	"time"
-
-	"git.eeqj.de/sneak/routewatch/internal/config"
-	"git.eeqj.de/sneak/routewatch/internal/database"
-	"git.eeqj.de/sneak/routewatch/internal/logger"
-	"git.eeqj.de/sneak/routewatch/internal/metrics"
-	"git.eeqj.de/sneak/routewatch/internal/server"
-	"git.eeqj.de/sneak/routewatch/internal/streamer"
-	"github.com/google/uuid"
-)
-
-// mockStore is a mock implementation of database.Store for testing
-type mockStore struct {
-	mu sync.Mutex
-
-	// Counters for tracking calls
-	ASNCount        int
-	PrefixCount     int
-	PeeringCount    int
-	RouteCount      int
-	WithdrawalCount int
-
-	// Track unique items
-	ASNs     map[int]*database.ASN
-	Prefixes map[string]*database.Prefix
-	Peerings map[string]bool // key is "from_to"
-	Routes   map[string]bool // key is "prefix_origin_peer"
-
-	// Track IP versions
-	IPv4Prefixes int
-	IPv6Prefixes int
-}
-
-// newMockStore creates a new mock store
-func newMockStore() *mockStore {
-	return &mockStore{
-		ASNs:     make(map[int]*database.ASN),
-		Prefixes: make(map[string]*database.Prefix),
-		Peerings: make(map[string]bool),
-		Routes:   make(map[string]bool),
-	}
-}
-
-// GetOrCreateASN mock implementation
-func (m *mockStore) GetOrCreateASN(number int, timestamp time.Time) (*database.ASN, error) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	if asn, exists := m.ASNs[number]; exists {
-		asn.LastSeen = timestamp
-
-		return asn, nil
-	}
-
-	asn := &database.ASN{
-		ASN:       number,
-		FirstSeen: timestamp,
-		LastSeen:  timestamp,
-	}
-	m.ASNs[number] = asn
-	m.ASNCount++
-
-	return asn, nil
-}
-
-// UpdatePrefixesBatch mock implementation
-func (m *mockStore) UpdatePrefixesBatch(prefixes map[string]time.Time) error {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	for prefix, timestamp := range prefixes {
-		if p, exists := m.Prefixes[prefix]; exists {
-			p.LastSeen = timestamp
-		} else {
-			const (
-				ipVersionV4 = 4
-				ipVersionV6 = 6
-			)
-
-			ipVersion := ipVersionV4
-			if strings.Contains(prefix, ":") {
-				ipVersion = ipVersionV6
-			}
-
-			m.Prefixes[prefix] = &database.Prefix{
-				ID:        uuid.New(),
-				Prefix:    prefix,
-				IPVersion: ipVersion,
-				FirstSeen: timestamp,
-				LastSeen:  timestamp,
-			}
-		}
-	}
-	return nil
-}
-
-// GetOrCreatePrefix mock implementation
-func (m *mockStore) GetOrCreatePrefix(prefix string, timestamp time.Time) (*database.Prefix, error) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	if p, exists := m.Prefixes[prefix]; exists {
-		p.LastSeen = timestamp
-
-		return p, nil
-	}
-
-	const (
-		ipVersionV4 = 4
-		ipVersionV6 = 6
-	)
-
-	ipVersion := ipVersionV4
-	if strings.Contains(prefix, ":") {
-		ipVersion = ipVersionV6
-	}
-
-	p := &database.Prefix{
-		ID:        uuid.New(),
-		Prefix:    prefix,
-		IPVersion: ipVersion,
-		FirstSeen: timestamp,
-		LastSeen:  timestamp,
-	}
-	m.Prefixes[prefix] = p
-	m.PrefixCount++
-
-	if ipVersion == ipVersionV4 {
-		m.IPv4Prefixes++
-	} else {
-		m.IPv6Prefixes++
-	}
-
-	return p, nil
-}
-
-// RecordAnnouncement mock implementation
-func (m *mockStore) RecordAnnouncement(_ *database.Announcement) error {
-	// Not tracking announcements in detail for now
-	return nil
-}
-
-// RecordPeering mock implementation
-func (m *mockStore) RecordPeering(asA, asB int, _ time.Time) error {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	// Normalize
-	if asA > asB {
-		asA, asB = asB, asA
-	}
-
-	key := fmt.Sprintf("%d_%d", asA, asB)
-	if !m.Peerings[key] {
-		m.Peerings[key] = true
-		m.PeeringCount++
-	}
-
-	return nil
-}
-
-// UpdatePeer mock implementation
-func (m *mockStore) UpdatePeer(peerIP string, peerASN int, messageType string, timestamp time.Time) error {
-	// Simple mock - just return nil
-	return nil
-}
-
-// Close mock implementation
-func (m *mockStore) Close() error {
-	return nil
-}
-
-// GetStats returns statistics about the mock store
-func (m *mockStore) GetStats() (database.Stats, error) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	return database.Stats{
-		ASNs:         len(m.ASNs),
-		Prefixes:     len(m.Prefixes),
-		IPv4Prefixes: m.IPv4Prefixes,
-		IPv6Prefixes: m.IPv6Prefixes,
-		Peerings:     m.PeeringCount,
-		Peers:        10, // Mock peer count
-	}, nil
-}
-
-// GetStatsContext returns statistics about the mock store with context support
-func (m *mockStore) GetStatsContext(ctx context.Context) (database.Stats, error) {
-	return m.GetStats()
-}
-
-// UpsertLiveRoute mock implementation
-func (m *mockStore) UpsertLiveRoute(route *database.LiveRoute) error {
-	// Simple mock - just return nil
-	return nil
-}
-
-// DeleteLiveRoute mock implementation
-func (m *mockStore) DeleteLiveRoute(prefix string, originASN int, peerIP string) error {
-	// Simple mock - just return nil
-	return nil
-}
-
-// GetPrefixDistribution mock implementation
-func (m *mockStore) GetPrefixDistribution() (ipv4 []database.PrefixDistribution, ipv6 []database.PrefixDistribution, err error) {
-	// Return empty distributions for now
-	return nil, nil, nil
-}
-
-// GetPrefixDistributionContext mock implementation with context support
-func (m *mockStore) GetPrefixDistributionContext(ctx context.Context) (ipv4 []database.PrefixDistribution, ipv6 []database.PrefixDistribution, err error) {
-	return m.GetPrefixDistribution()
-}
-
-// GetLiveRouteCounts mock implementation
-func (m *mockStore) GetLiveRouteCounts() (ipv4Count, ipv6Count int, err error) {
-	// Return mock counts
-	return m.RouteCount / 2, m.RouteCount / 2, nil
-}
-
-// GetLiveRouteCountsContext mock implementation with context support
-func (m *mockStore) GetLiveRouteCountsContext(ctx context.Context) (ipv4Count, ipv6Count int, err error) {
-	return m.GetLiveRouteCounts()
-}
-
-// GetASDetails mock implementation
-func (m *mockStore) GetASDetails(asn int) (*database.ASN, []database.LiveRoute, error) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	// Check if ASN exists
-	if asnInfo, exists := m.ASNs[asn]; exists {
-		// Return empty prefixes for now
-		return asnInfo, []database.LiveRoute{}, nil
-	}
-
-	return nil, nil, database.ErrNoRoute
-}
-
-// GetASDetailsContext mock implementation with context support
-func (m *mockStore) GetASDetailsContext(ctx context.Context, asn int) (*database.ASN, []database.LiveRoute, error) {
-	return m.GetASDetails(asn)
-}
-
-// GetPrefixDetails mock implementation
-func (m *mockStore) GetPrefixDetails(prefix string) ([]database.LiveRoute, error) {
-	// Return empty routes for now
-	return []database.LiveRoute{}, nil
-}
-
-// GetPrefixDetailsContext mock implementation with context support
-func (m *mockStore) GetPrefixDetailsContext(ctx context.Context, prefix string) ([]database.LiveRoute, error) {
-	return m.GetPrefixDetails(prefix)
-}
-
-func (m *mockStore) GetRandomPrefixesByLength(maskLength, ipVersion, limit int) ([]database.LiveRoute, error) {
-	// Return empty routes for now
-	return []database.LiveRoute{}, nil
-}
-
-// GetRandomPrefixesByLengthContext mock implementation with context support
-func (m *mockStore) GetRandomPrefixesByLengthContext(ctx context.Context, maskLength, ipVersion, limit int) ([]database.LiveRoute, error) {
-	return m.GetRandomPrefixesByLength(maskLength, ipVersion, limit)
-}
-
-// GetASPeers mock implementation
-func (m *mockStore) GetASPeers(asn int) ([]database.ASPeer, error) {
-	// Return empty peers for now
-	return []database.ASPeer{}, nil
-}
-
-// GetASPeersContext mock implementation with context support
-func (m *mockStore) GetASPeersContext(ctx context.Context, asn int) ([]database.ASPeer, error) {
-	return m.GetASPeers(asn)
-}
-
-// GetIPInfo mock implementation
-func (m *mockStore) GetIPInfo(ip string) (*database.IPInfo, error) {
-	return m.GetIPInfoContext(context.Background(), ip)
-}
-
-// GetIPInfoContext mock implementation with context support
-func (m *mockStore) GetIPInfoContext(ctx context.Context, ip string) (*database.IPInfo, error) {
-	now := time.Now()
-	return &database.IPInfo{
-		IP:          ip,
-		Netblock:    "8.8.8.0/24",
-		MaskLength:  24,
-		IPVersion:   4,
-		NumPeers:    3,
-		ASN:         15169,
-		Handle:      "GOOGLE",
-		Description: "Google LLC",
-		CountryCode: "US",
-		FirstSeen:   now.Add(-24 * time.Hour),
-		LastSeen:    now,
-	}, nil
-}
-
-// GetNextStaleASN mock implementation
-func (m *mockStore) GetNextStaleASN(ctx context.Context, staleThreshold time.Duration) (int, error) {
-	return 0, database.ErrNoStaleASN
-}
-
-// UpdateASNWHOIS mock implementation
-func (m *mockStore) UpdateASNWHOIS(ctx context.Context, update *database.ASNWHOISUpdate) error {
-	return nil
-}
-
-// GetWHOISStats mock implementation
-func (m *mockStore) GetWHOISStats(ctx context.Context, staleThreshold time.Duration) (*database.WHOISStats, error) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	return &database.WHOISStats{
-		TotalASNs:    len(m.ASNs),
-		FreshASNs:    0,
-		StaleASNs:    0,
-		NeverFetched: len(m.ASNs),
-	}, nil
-}
-
-// UpsertLiveRouteBatch mock implementation
-func (m *mockStore) UpsertLiveRouteBatch(routes []*database.LiveRoute) error {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	for _, route := range routes {
-		// Track prefix
-		if _, exists := m.Prefixes[route.Prefix]; !exists {
-			m.Prefixes[route.Prefix] = &database.Prefix{
-				ID:        uuid.New(),
-				Prefix:    route.Prefix,
-				IPVersion: route.IPVersion,
-				FirstSeen: route.LastUpdated,
-				LastSeen:  route.LastUpdated,
-			}
-			m.PrefixCount++
-			if route.IPVersion == 4 {
-				m.IPv4Prefixes++
-			} else {
-				m.IPv6Prefixes++
-			}
-		}
-		m.RouteCount++
-	}
-	return nil
-}
-
-// DeleteLiveRouteBatch mock implementation
-func (m *mockStore) DeleteLiveRouteBatch(deletions []database.LiveRouteDeletion) error {
-	// Simple mock - just return nil
-	return nil
-}
-
-// GetOrCreateASNBatch mock implementation
-func (m *mockStore) GetOrCreateASNBatch(asns map[int]time.Time) error {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	for number, timestamp := range asns {
-		if _, exists := m.ASNs[number]; !exists {
-			m.ASNs[number] = &database.ASN{
-				ASN:       number,
-				FirstSeen: timestamp,
-				LastSeen:  timestamp,
-			}
-			m.ASNCount++
-		}
-	}
-	return nil
-}
-
-// UpdatePeerBatch mock implementation
-func (m *mockStore) UpdatePeerBatch(peers map[string]database.PeerUpdate) error {
-	// Simple mock - just return nil
-	return nil
-}
-
-// Vacuum mock implementation
-func (m *mockStore) Vacuum(ctx context.Context) error {
-	return nil
-}
-
-// Analyze mock implementation
-func (m *mockStore) Analyze(ctx context.Context) error {
-	return nil
-}
-
-// Checkpoint mock implementation
-func (m *mockStore) Checkpoint(ctx context.Context) error {
-	return nil
-}
-
-// Ping mock implementation
-func (m *mockStore) Ping(ctx context.Context) error {
-	return nil
-}
-
-func TestRouteWatchLiveFeed(t *testing.T) {
-	if testing.Short() {
-		t.Skip("skipping live RIPE RIS network feed test in short mode; run without -short to include it")
-	}
-
-	// Create mock database
-	mockDB := newMockStore()
-	defer mockDB.Close()
-
-	logger := logger.New()
-
-	// Create metrics tracker
-	metricsTracker := metrics.New()
-
-	// Create streamer
-	s := streamer.New(logger, metricsTracker)
-
-	// Create test config with empty state dir (no snapshot loading)
-	cfg := &config.Config{
-		StateDir:                    "",
-		MaxRuntime:                  5 * time.Second,
-		EnableBatchedDatabaseWrites: true,
-	}
-
-	// Create server
-	srv := server.New(mockDB, s, logger, cfg)
-
-	// Create RouteWatch with 5 second limit
-	deps := Dependencies{
-		DB:       mockDB,
-		Streamer: s,
-		Server:   srv,
-		Logger:   logger,
-		Config:   cfg,
-	}
-	rw := New(deps)
-
-	// Run with context
-	ctx := context.Background()
-	go func() {
-		_ = rw.Run(ctx)
-	}()
-
-	// Wait for the configured duration
-	time.Sleep(5 * time.Second)
-
-	// Force peering processing for test
-	if rw.peeringHandler != nil {
-		rw.peeringHandler.ProcessPeeringsNow()
-	}
-
-	// Get statistics
-	stats, err := mockDB.GetStats()
-	if err != nil {
-		t.Fatalf("Failed to get stats: %v", err)
-	}
-
-	if stats.ASNs == 0 {
-		t.Error("Expected to receive some ASNs from live feed")
-	}
-	t.Logf("Received %d unique ASNs in 5 seconds", stats.ASNs)
-
-	if stats.Prefixes == 0 {
-		t.Error("Expected to receive some prefixes from live feed")
-	}
-	t.Logf("Received %d unique prefixes (%d IPv4, %d IPv6) in 5 seconds", stats.Prefixes, stats.IPv4Prefixes, stats.IPv6Prefixes)
-
-	if stats.Peerings == 0 {
-		t.Error("Expected to receive some peerings from live feed")
-	}
-	t.Logf("Recorded %d AS peering relationships in 5 seconds", stats.Peerings)
-
-}

internal/routewatch/app_test.go

@@ -1,3 +1,3 @@
-package routewatch
+package routewatch_test
 
-// Tests for routewatch package are in app_integration_test.go
+// Tests for routewatch package are in app_integration_internal_test.go

internal/routewatch/ashandler.go

@@ -1,6 +1,8 @@
 package routewatch
 
 import (
+	"context"
+	"slices"
 	"sync"
 	"time"
 
@@ -19,6 +21,12 @@ const (
 	// asnBatchSize is the number of ASN operations to batch together
 	asnBatchSize = 30000
 
+	// asnBatchMaxSize is the most ASN operations the batch holds while the
+	// flush goroutine writes the previous one: ten batches, about 10 MiB
+	// (README "Memory"). At this size HandleMessage waits until the flush
+	// goroutine takes the batch.
+	asnBatchMaxSize = 300000
+
 	// asnBatchTimeout is the maximum time to wait before flushing a batch
 	// DO NOT reduce this timeout - larger batches are more efficient
 	asnBatchTimeout = 2 * time.Second
@@ -32,12 +40,23 @@ type ASHandler struct {
 	db     database.Store
 	logger *logger.Logger
 
-	// Batching
+	// Batching. mu guards batch and writing only: the flush goroutine takes
+	// the batch under mu and writes it after releasing mu, so HandleMessage
+	// waits for a database write only while batch is at asnBatchMaxSize.
 	mu    sync.Mutex
 	batch []asnOp
-	lastFlush time.Time
+	// writing is how many operations the flush goroutine took from batch
-	stopCh    chan struct{}
+	// and has not finished writing.
-	wg        sync.WaitGroup
+	writing int
+	// batchTaken is signalled when the flush goroutine takes batch; its L is
+	// &mu.
+	batchTaken  sync.Cond
+	batchFullCh chan struct{} // wakes the flush goroutine when batch is full
+	lastFlush   time.Time     // used only by the flush goroutine
+	// stopCh hands Stop's context to the flush goroutine. It holds one, so
+	// Stop does not wait for a write in progress to hand it over.
+	stopCh chan context.Context
+	done   chan struct{} // closed when the flush goroutine returns
 }
 
 type asnOp struct {
@@ -55,12 +74,14 @@ func NewASHandler(db database.Store, logger *logger.Logger) *ASHandler {
 		db:          db,
 		logger:      logger,
 		batch:       make([]asnOp, 0, asnBatchSize),
+		batchFullCh: make(chan struct{}, 1),
 		lastFlush:   time.Now(),
-		stopCh:    make(chan struct{}),
+		stopCh:      make(chan context.Context, 1),
+		done:        make(chan struct{}),
 	}
+	h.batchTaken.L = &h.mu
 
 	// Start the flush timer goroutine
-	h.wg.Add(1)
 	go h.flushLoop()
 
 	return h
@@ -71,7 +92,7 @@ func NewASHandler(db database.Store, logger *logger.Logger) *ASHandler {
 // the AS path information needed to track autonomous systems.
 func (h *ASHandler) WantsMessage(messageType string) bool {
 	// We only care about UPDATE messages for the database
-	return messageType == "UPDATE"
+	return messageType == ristypes.MessageTypeUpdate
 }
 
 // QueueCapacity returns the recommended message queue size for this handler.
@@ -84,22 +105,25 @@ func (h *ASHandler) QueueCapacity() int {
 
 // HandleMessage processes a RIS Live BGP message by extracting all ASNs from
 // the AS path and queuing them for batch insertion into the database. The
-// origin ASN (last element in the path) and all transit ASNs are recorded
+// origin ASN and all transit ASNs are recorded with their associated
-// with their associated timestamps. The batch is automatically flushed when
+// timestamps. When the batch reaches the configured size threshold, the flush
-// it reaches the configured size threshold.
+// goroutine is woken to write it.
 func (h *ASHandler) HandleMessage(msg *ristypes.RISMessage) {
 	// Use the pre-parsed timestamp
 	timestamp := msg.ParsedTimestamp
 
-	// Get origin ASN from path (last element)
+	originASN := msg.Path.Origin()
-	var originASN int
-	if len(msg.Path) > 0 {
-		originASN = msg.Path[len(msg.Path)-1]
-	}
 
 	h.mu.Lock()
 	defer h.mu.Unlock()
 
+	// While the batch is at its maximum size, wait until the flush goroutine
+	// takes it. Meanwhile the handler takes nothing off its queue, which
+	// fills, and the streamer drops messages for it.
+	for len(h.batch) >= asnBatchMaxSize {
+		h.batchTaken.Wait()
+	}
+
 	// Queue origin ASN operation
 	if originASN > 0 {
 		h.batch = append(h.batch, asnOp{
@@ -108,8 +132,8 @@ func (h *ASHandler) HandleMessage(msg *ristypes.RISMessage) {
 		})
 	}
 
-	// Also track all ASNs in the path
+	// Also track all ASNs in the path, the members of an AS set included
-	for _, asn := range msg.Path {
+	for _, asn := range slices.Concat(msg.Path.Sequence, msg.Path.Set) {
 		if asn > 0 {
 			h.batch = append(h.batch, asnOp{
 				number:    asn,
@@ -118,67 +142,113 @@ func (h *ASHandler) HandleMessage(msg *ristypes.RISMessage) {
 		}
 	}
 
-	// Check if we need to flush
+	// A full batch wakes the flush goroutine. If it is writing, the wake-up
+	// waits in batchFullCh and the batch keeps growing until the write ends
+	// or the batch reaches asnBatchMaxSize.
 	if len(h.batch) >= asnBatchSize {
-		h.flushBatchLocked()
+		select {
+		case h.batchFullCh <- struct{}{}:
+		default: // already woken
+		}
 	}
 }
 
-// flushLoop runs in a goroutine and periodically flushes batches
+// Stop stops the flush goroutine, which writes the operations the handler holds
+// before it returns, and waits for it until ctx ends. Once ctx has ended the
+// flush goroutine skips that last write, and a write still running is lost
+// when the process exits, so Stop logs how many operations are not written.
+func (h *ASHandler) Stop(ctx context.Context) {
+	h.stopCh <- ctx
+
+	select {
+	case <-h.done:
+	case <-ctx.Done():
+	}
+
+	h.mu.Lock()
+	unwritten := h.writing + len(h.batch)
+	h.mu.Unlock()
+
+	if unwritten > 0 {
+		h.logger.Error("Stop deadline passed before the AS handler wrote "+
+			"its updates", "updates", unwritten)
+	}
+}
+
+// flushLoop runs in a goroutine, the only one that writes the batch: it
+// flushes when the batch is full, periodically, and once more on Stop.
 func (h *ASHandler) flushLoop() {
-	defer h.wg.Done()
+	defer close(h.done)
+
 	ticker := time.NewTicker(asnBatchTimeout)
 	defer ticker.Stop()
 
 	for {
 		select {
 		case <-ticker.C:
-			h.mu.Lock()
 			if time.Since(h.lastFlush) >= asnBatchTimeout {
-				h.flushBatchLocked()
+				h.flushBatch()
 			}
-			h.mu.Unlock()
+		case <-h.batchFullCh:
-		case <-h.stopCh:
+			h.flushBatch()
-			// Final flush
+		case ctx := <-h.stopCh:
+			// Final flush, unless the stop deadline has passed
+			if ctx.Err() == nil {
+				h.flushBatch()
+			}
+
+			return
+		}
+	}
+}
+
+// takeBatch returns the pending operations and leaves an empty batch in their
+// place for HandleMessage to fill. They count in writing until flushBatch has
+// written them.
+func (h *ASHandler) takeBatch() []asnOp {
 	h.mu.Lock()
-			h.flushBatchLocked()
+	defer h.mu.Unlock()
-			h.mu.Unlock()
 
-			return
-		}
-	}
-}
-
-// flushBatchLocked flushes the ASN batch to the database (must be called with mutex held)
-func (h *ASHandler) flushBatchLocked() {
 	if len(h.batch) == 0 {
+		return nil
+	}
+
+	batch := h.batch
+	h.writing = len(batch)
+	h.batch = make([]asnOp, 0, asnBatchSize)
+	h.batchTaken.Broadcast()
+
+	return batch
+}
+
+// flushBatch takes the pending batch and writes it to the database, without
+// holding the mutex during the write.
+func (h *ASHandler) flushBatch() {
+	batch := h.takeBatch()
+	if len(batch) == 0 {
 		return
 	}
 
-	// Process ASNs first (deduped)
+	// The earliest and the latest time the batch saw each AS. The batch holds
-	asnMap := make(map[int]time.Time)
+	// several updates for an AS, from many peers, so a new AS is stored first
-	for _, op := range h.batch {
+	// seen at the earliest of them and last seen at the latest.
-		if existing, ok := asnMap[op.number]; !ok || op.timestamp.After(existing) {
+	asns := make(map[int]database.SeenTimes)
-			asnMap[op.number] = op.timestamp
+	for _, op := range batch {
-		}
+		seen := asns[op.number]
+		seen.Add(op.timestamp)
+		asns[op.number] = seen
 	}
 
 	// Process all ASNs in a single batch transaction
-	if err := h.db.GetOrCreateASNBatch(asnMap); err != nil {
+	err := h.db.GetOrCreateASNBatch(asns)
-		h.logger.Error("Failed to process ASN batch", "error", err, "count", len(asnMap))
+	if err != nil {
+		h.logger.Error("Failed to process ASN batch",
+			"error", err, "count", len(asns))
 	}
 
-	// Clear batch
+	h.mu.Lock()
-	h.batch = h.batch[:0]
+	h.writing = 0
+	h.mu.Unlock()
+
 	h.lastFlush = time.Now()
 }
-
-// Stop gracefully shuts down the ASHandler by signaling the background flush
-// goroutine to terminate and waiting for it to complete. Any pending ASN
-// operations in the current batch are flushed to the database before Stop
-// returns. This method should be called during application shutdown to ensure
-// no data is lost.
-func (h *ASHandler) Stop() {
-	close(h.stopCh)
-	h.wg.Wait()
-}

internal/routewatch/ashandler_internal_test.go

@@ -0,0 +1,68 @@
+package routewatch
+
+import (
+	"maps"
+	"slices"
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+)
+
+// TestASHandlerStoresTheEarliestAndLatestTimesOfANewAS handles two updates
+// whose path carries an AS the database does not hold yet, the second 5
+// seconds after the first, in one batch into a real database. The AS is stored
+// first seen at the time of the first update and last seen at the time of the
+// second.
+func TestASHandlerStoresTheEarliestAndLatestTimesOfANewAS(t *testing.T) {
+	t.Parallel()
+
+	db := newTestDatabase(t)
+
+	// Built without NewASHandler's flush goroutine, so the test flushes itself.
+	h := &ASHandler{db: db, logger: logger.New()}
+
+	first := peerAnnouncement(testPeerIP, testV4Prefix)
+	second := peerAnnouncement(testPeerIP, testV4Prefix)
+	second.ParsedTimestamp = first.ParsedTimestamp.Add(secondUpdateAfter)
+
+	h.HandleMessage(first)
+	h.HandleMessage(second)
+	h.flushBatch()
+
+	asn, _, err := db.GetASDetails(testASNA)
+	if err != nil {
+		t.Fatalf("GetASDetails(%d): %v", testASNA, err)
+	}
+
+	if !asn.FirstSeen.Equal(first.ParsedTimestamp) {
+		t.Errorf("first_seen of AS%d = %s, want %s", testASNA,
+			asn.FirstSeen, first.ParsedTimestamp)
+	}
+
+	if !asn.LastSeen.Equal(second.ParsedTimestamp) {
+		t.Errorf("last_seen of AS%d = %s, want %s", testASNA,
+			asn.LastSeen, second.ParsedTimestamp)
+	}
+}
+
+// TestASHandlerRecordsNothingFromAPathWithASZero handles an update whose AS
+// path holds AS 0 between two other ASes. RFC 7607 makes such a path
+// malformed, so neither AS is recorded.
+func TestASHandlerRecordsNothingFromAPathWithASZero(t *testing.T) {
+	t.Parallel()
+
+	store := newMockStore()
+
+	// Built without NewASHandler's flush goroutine, so the test flushes itself.
+	h := &ASHandler{db: store, logger: logger.New()}
+
+	msg := peerAnnouncement(testPeerIP, testV4Prefix)
+	msg.Path = decodePath(t, testASNA, 0, testASNB)
+	h.HandleMessage(msg)
+	h.flushBatch()
+
+	if len(store.ASNs) > 0 {
+		t.Errorf("ASes recorded = %v, want none",
+			slices.Sorted(maps.Keys(store.ASNs)))
+	}
+}

internal/routewatch/asnfetcher.go

@@ -1,8 +1,8 @@
-// Package routewatch contains the ASN WHOIS fetcher for background updates.
 package routewatch
 
 import (
 	"context"
+	"errors"
 	"log/slog"
 	"sync"
 	"time"
@@ -21,10 +21,16 @@ const (
 	minInterval = 1 * time.Second
 
 	// maxInterval is the maximum interval after failures (backoff cap).
-	maxInterval = 5 * time.Minute
+	maxInterval = time.Minute
 
-	// backoffMultiplier is how much to multiply interval on failure.
+	// A failure multiplies the interval by backoffNumerator divided by
-	backoffMultiplier = 2
+	// backoffDenominator, 1.5, up to maxInterval.
+	backoffNumerator   = 3
+	backoffDenominator = 2
+
+	// A success divides the interval by successDivisor, down to
+	// minInterval.
+	successDivisor = 2
 
 	// whoisStaleThreshold is how old WHOIS data can be before refresh.
 	whoisStaleThreshold = 30 * 24 * time.Hour // 30 days
@@ -43,7 +49,7 @@ type ASNFetcher struct {
 	logger         *slog.Logger
 	immediateQueue chan int
 	stopCh         chan struct{}
-	wg             sync.WaitGroup
+	done           chan struct{} // closed when the fetcher goroutine returns
 
 	// fetchMu ensures only one fetch runs at a time
 	fetchMu sync.Mutex
@@ -53,7 +59,8 @@ type ASNFetcher struct {
 	currentInterval  time.Duration
 	consecutiveFails int
 
-	// hourly stats tracking
+	// hourly stats tracking; now is the clock, time.Now except in tests
+	now          func() time.Time
 	statsMu      sync.Mutex
 	successTimes []time.Time
 	errorTimes   []time.Time
@@ -61,22 +68,27 @@ type ASNFetcher struct {
 
 // NewASNFetcher creates a new ASN fetcher.
 func NewASNFetcher(db database.Store, logger *slog.Logger) *ASNFetcher {
+	logger = logger.With("component", "asn_fetcher")
+
 	return &ASNFetcher{
 		db:              db,
-		whoisClient:     whois.NewClient(),
+		whoisClient:     whois.NewClient(logger),
-		logger:          logger.With("component", "asn_fetcher"),
+		logger:          logger,
 		immediateQueue:  make(chan int, immediateQueueSize),
 		stopCh:          make(chan struct{}),
+		done:            make(chan struct{}),
 		currentInterval: baseInterval,
+		now:             time.Now,
 		successTimes:    make([]time.Time, 0),
 		errorTimes:      make([]time.Time, 0),
 	}
 }
 
-// Start begins the background ASN fetcher goroutine.
+// Start begins the background ASN fetcher goroutine. It runs until Stop:
-func (f *ASNFetcher) Start() {
+// the end of ctx does not stop it.
-	f.wg.Add(1)
+func (f *ASNFetcher) Start(ctx context.Context) {
-	go f.run()
+	go f.run(context.WithoutCancel(ctx))
+
 	f.logger.Info("ASN fetcher started",
 		"base_interval", baseInterval,
 		"min_interval", minInterval,
@@ -84,11 +96,17 @@ func (f *ASNFetcher) Start() {
 	)
 }
 
-// Stop gracefully shuts down the fetcher.
+// Stop stops the fetcher and waits, until ctx ends, for a lookup in progress to
-func (f *ASNFetcher) Stop() {
+// finish.
+func (f *ASNFetcher) Stop(ctx context.Context) {
 	close(f.stopCh)
-	f.wg.Wait()
+
+	select {
+	case <-f.done:
 		f.logger.Info("ASN fetcher stopped")
+	case <-ctx.Done():
+		f.logger.Warn("Stop deadline passed before the ASN fetcher stopped")
+	}
 }
 
 // QueueImmediate queues an ASN for immediate WHOIS lookup.
@@ -113,7 +131,7 @@ func (f *ASNFetcher) GetStats() server.ASNFetcherStats {
 	f.intervalMu.Unlock()
 
 	// Prune old entries and count
-	cutoff := time.Now().Add(-statsWindow)
+	cutoff := f.now().Add(-statsWindow)
 	f.successTimes = pruneOldTimes(f.successTimes, cutoff)
 	f.errorTimes = pruneOldTimes(f.errorTimes, cutoff)
 
@@ -125,16 +143,15 @@ func (f *ASNFetcher) GetStats() server.ASNFetcherStats {
 	}
 }
 
-// pruneOldTimes removes times older than cutoff and returns the pruned slice.
+// pruneOldTimes drops the times at or before cutoff from the front of times,
+// which holds times in the order they were recorded, and returns the rest. It
+// copies nothing, so it is cheap enough to run on every append.
 func pruneOldTimes(times []time.Time, cutoff time.Time) []time.Time {
-	result := make([]time.Time, 0, len(times))
+	for len(times) > 0 && !times[0].After(cutoff) {
-	for _, t := range times {
+		times = times[1:]
-		if t.After(cutoff) {
-			result = append(result, t)
-		}
 	}
 
-	return result
+	return times
 }
 
 // getInterval returns the current fetch interval.
@@ -151,10 +168,7 @@ func (f *ASNFetcher) recordSuccess() {
 	f.consecutiveFails = 0
 
 	// Decrease interval by half, but not below minimum
-	newInterval := f.currentInterval / backoffMultiplier
+	newInterval := max(f.currentInterval/successDivisor, minInterval)
-	if newInterval < minInterval {
-		newInterval = minInterval
-	}
 
 	if newInterval != f.currentInterval {
 		f.logger.Debug("Decreased fetch interval",
@@ -165,9 +179,11 @@ func (f *ASNFetcher) recordSuccess() {
 	}
 	f.intervalMu.Unlock()
 
-	// Record success time for stats
+	// Record success time for stats, dropping those older than statsWindow
 	f.statsMu.Lock()
-	f.successTimes = append(f.successTimes, time.Now())
+	now := f.now()
+	f.successTimes = pruneOldTimes(f.successTimes, now.Add(-statsWindow))
+	f.successTimes = append(f.successTimes, now)
 	f.statsMu.Unlock()
 }
 
@@ -176,11 +192,9 @@ func (f *ASNFetcher) recordFailure() {
 	f.intervalMu.Lock()
 	f.consecutiveFails++
 
-	// Exponential backoff: multiply by 2, capped at max
+	// Exponential backoff: multiply by 1.5, capped at max
-	newInterval := f.currentInterval * backoffMultiplier
+	newInterval := min(
-	if newInterval > maxInterval {
+		f.currentInterval*backoffNumerator/backoffDenominator, maxInterval)
-		newInterval = maxInterval
-	}
 
 	if newInterval != f.currentInterval {
 		f.logger.Debug("Increased fetch interval due to failure",
@@ -192,15 +206,17 @@ func (f *ASNFetcher) recordFailure() {
 	}
 	f.intervalMu.Unlock()
 
-	// Record error time for stats
+	// Record error time for stats, dropping those older than statsWindow
 	f.statsMu.Lock()
-	f.errorTimes = append(f.errorTimes, time.Now())
+	now := f.now()
+	f.errorTimes = pruneOldTimes(f.errorTimes, now.Add(-statsWindow))
+	f.errorTimes = append(f.errorTimes, now)
 	f.statsMu.Unlock()
 }
 
 // run is the main background loop.
-func (f *ASNFetcher) run() {
+func (f *ASNFetcher) run(ctx context.Context) {
-	defer f.wg.Done()
+	defer close(f.done)
 
 	timer := time.NewTimer(f.getInterval())
 	defer timer.Stop()
@@ -212,13 +228,13 @@ func (f *ASNFetcher) run() {
 
 		case asn := <-f.immediateQueue:
 			// Process immediate request (respects lock)
-			f.tryFetch(asn)
+			f.tryFetch(ctx, asn)
 			// Reset timer after immediate fetch
 			timer.Reset(f.getInterval())
 
 		case <-timer.C:
 			// Background fetch of stale/missing ASN
-			f.fetchNextStale()
+			f.fetchNextStale(ctx)
 			// Reset timer with potentially updated interval
 			timer.Reset(f.getInterval())
 		}
@@ -227,7 +243,7 @@ func (f *ASNFetcher) run() {
 
 // tryFetch attempts to fetch and update an ASN, respecting the fetch lock.
 // Returns true if fetch was successful.
-func (f *ASNFetcher) tryFetch(asn int) bool {
+func (f *ASNFetcher) tryFetch(ctx context.Context, asn int) bool {
 	// Try to acquire lock, skip if another fetch is running
 	if !f.fetchMu.TryLock() {
 		f.logger.Debug("Skipping fetch, another fetch in progress", "asn", asn)
@@ -236,11 +252,11 @@ func (f *ASNFetcher) tryFetch(asn int) bool {
 	}
 	defer f.fetchMu.Unlock()
 
-	return f.fetchAndUpdate(asn)
+	return f.fetchAndUpdate(ctx, asn)
 }
 
 // fetchNextStale finds and fetches the next ASN needing WHOIS data.
-func (f *ASNFetcher) fetchNextStale() {
+func (f *ASNFetcher) fetchNextStale(ctx context.Context) {
 	// Try to acquire lock, skip if another fetch is running
 	if !f.fetchMu.TryLock() {
 		f.logger.Debug("Skipping stale fetch, another fetch in progress")
@@ -249,12 +265,12 @@ func (f *ASNFetcher) fetchNextStale() {
 	}
 	defer f.fetchMu.Unlock()
 
-	ctx, cancel := context.WithTimeout(context.Background(), time.Minute)
+	lookupCtx, cancel := context.WithTimeout(ctx, time.Minute)
 	defer cancel()
 
-	asn, err := f.db.GetNextStaleASN(ctx, whoisStaleThreshold)
+	asn, err := f.db.GetNextStaleASN(lookupCtx, whoisStaleThreshold)
 	if err != nil {
-		if err != database.ErrNoStaleASN {
+		if !errors.Is(err, database.ErrNoStaleASN) {
 			f.logger.Error("Failed to get stale ASN", "error", err)
 			f.recordFailure()
 		}
@@ -263,18 +279,34 @@ func (f *ASNFetcher) fetchNextStale() {
 		return
 	}
 
-	f.fetchAndUpdate(asn)
+	f.fetchAndUpdate(ctx, asn)
 }
 
 // fetchAndUpdate performs a WHOIS lookup and updates the database.
-// Returns true if successful.
+// Returns true if successful. A failed lookup leaves the AS's row as it was,
-func (f *ASNFetcher) fetchAndUpdate(asn int) bool {
+// so the AS stays stale. An AS IANA refers to no RIR is stored with no
-	ctx, cancel := context.WithTimeout(context.Background(), time.Minute)
+// WHOIS data, so it is not looked up again until it is stale; that counts
+// as neither a success nor a failure.
+func (f *ASNFetcher) fetchAndUpdate(ctx context.Context, asn int) bool {
+	ctx, cancel := context.WithTimeout(ctx, time.Minute)
 	defer cancel()
 
 	f.logger.Info("Fetching WHOIS data", "asn", asn)
 
 	info, err := f.whoisClient.LookupASN(ctx, asn)
+	if errors.Is(err, whois.ErrNoReferral) {
+		f.logger.Info("IANA refers the AS to no RIR", "asn", asn)
+
+		err = f.db.UpdateASNWHOIS(ctx, &database.ASNWHOISUpdate{ASN: asn})
+		if err != nil {
+			f.logger.Error("Failed to update ASN WHOIS data",
+				"asn", asn, "error", err)
+			f.recordFailure()
+		}
+
+		return false
+	}
+
 	if err != nil {
 		f.logger.Error("WHOIS lookup failed", "asn", asn, "error", err)
 		f.recordFailure()

internal/routewatch/asnfetcher_internal_test.go

@@ -0,0 +1,278 @@
+package routewatch
+
+import (
+	"bufio"
+	"context"
+	"errors"
+	"io"
+	"net"
+	"sync"
+	"syscall"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+)
+
+// The WHOIS servers a lookup asks: IANA first, then the RIR IANA refers it
+// to.
+const (
+	ianaWHOISServer = "whois.iana.org:43"
+	ripeWHOISServer = "whois.ripe.net:43"
+)
+
+// ianaRefersToRIPE is IANA's answer for an AS in a block it gave to RIPE.
+const ianaRefersToRIPE = `refer:        whois.ripe.net
+
+as-block:     64496-64511
+organisation: Assigned by RIPE NCC
+changed:      2002-09-01
+source:       IANA
+`
+
+// ripeAccessDenied is RIPE's answer to a host it no longer serves.
+const ripeAccessDenied = `% This is the RIPE Database query service.
+%ERROR:201: access denied for 192.0.2.1
+`
+
+// ianaPrivateUseBlock is IANA's answer for a private-use AS: its block and
+// no referral.
+const ianaPrivateUseBlock = `as-block:     64512-65534
+organisation: Reserved for Private Use
+source:       IANA
+`
+
+// newFetcherWithWHOISAnswers returns an ASN fetcher on a new database that
+// holds testASNA, never looked up. The fetcher's WHOIS client connects to a
+// fake server for each address in answers, which reads the query and sends
+// answers[address]; connecting to any other address is refused.
+func newFetcherWithWHOISAnswers(
+	t *testing.T, answers map[string]string,
+) (*ASNFetcher, *database.Database) {
+	t.Helper()
+
+	db := newTestDatabase(t)
+
+	_, err := db.GetOrCreateASN(testASNA, time.Now())
+	if err != nil {
+		t.Fatalf("GetOrCreateASN: %v", err)
+	}
+
+	var servers sync.WaitGroup
+
+	t.Cleanup(servers.Wait)
+
+	f := NewASNFetcher(db, logger.New().Logger)
+	f.whoisClient.DialContext = func(
+		_ context.Context, _, address string,
+	) (net.Conn, error) {
+		answer, ok := answers[address]
+		if !ok {
+			return nil, syscall.ECONNREFUSED
+		}
+
+		client, server := net.Pipe()
+
+		servers.Go(func() {
+			defer func() { _ = server.Close() }()
+
+			_, _ = bufio.NewReader(server).ReadString('\n')
+			_, _ = io.WriteString(server, answer)
+		})
+
+		return client, nil
+	}
+
+	return f, db
+}
+
+// TestFetchAndUpdateKeepsTheRowWhenTheLookupFails looks up an AS IANA refers
+// to RIPE when RIPE's server refuses the connection, and when it answers only
+// with an error. The AS must stay never looked up, and the fetcher must back
+// off.
+func TestFetchAndUpdateKeepsTheRowWhenTheLookupFails(t *testing.T) {
+	t.Parallel()
+
+	for name, answers := range map[string]map[string]string{
+		"connection refused": {ianaWHOISServer: ianaRefersToRIPE},
+		"only an error": {
+			ianaWHOISServer: ianaRefersToRIPE,
+			ripeWHOISServer: ripeAccessDenied,
+		},
+	} {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			f, db := newFetcherWithWHOISAnswers(t, answers)
+
+			if f.fetchAndUpdate(t.Context(), testASNA) {
+				t.Error("fetchAndUpdate reported success")
+			}
+
+			stats, err := db.GetWHOISStats(t.Context(), whoisStaleThreshold)
+			if err != nil {
+				t.Fatalf("GetWHOISStats: %v", err)
+			}
+
+			if stats.NeverFetched != 1 {
+				t.Errorf("never looked up = %d, want 1: the failed lookup "+
+					"was stored", stats.NeverFetched)
+			}
+
+			want := baseInterval * backoffNumerator / backoffDenominator
+
+			got := f.GetStats()
+			if got.CurrentInterval != want || got.ConsecutiveFails != 1 {
+				t.Errorf("interval %v after %d failures, want %v after 1",
+					got.CurrentInterval, got.ConsecutiveFails, want)
+			}
+		})
+	}
+}
+
+// TestRecordStepsTheInterval has the fetcher fail from its base interval of
+// 15 seconds, then succeed. Each failure makes the interval 1.5 times as
+// long, up to one minute, and each success halves it, down to one second.
+func TestRecordStepsTheInterval(t *testing.T) {
+	t.Parallel()
+
+	f := NewASNFetcher(newMockStore(), logger.New().Logger)
+
+	for i, want := range []time.Duration{
+		22500 * time.Millisecond,
+		33750 * time.Millisecond,
+		50625 * time.Millisecond,
+		time.Minute,
+		time.Minute,
+	} {
+		f.recordFailure()
+
+		if got := f.getInterval(); got != want {
+			t.Errorf("interval after %d failures = %v, want %v",
+				i+1, got, want)
+		}
+	}
+
+	for i, want := range []time.Duration{
+		30 * time.Second,
+		15 * time.Second,
+		7500 * time.Millisecond,
+		3750 * time.Millisecond,
+		1875 * time.Millisecond,
+		time.Second,
+		time.Second,
+	} {
+		f.recordSuccess()
+
+		if got := f.getInterval(); got != want {
+			t.Errorf("interval after %d successes = %v, want %v",
+				i+1, got, want)
+		}
+	}
+}
+
+// TestFetchAndUpdateMarksAnASWithoutReferral looks up a private-use AS, which
+// IANA refers to no RIR. The AS must not be looked up again until it is
+// stale, and the lookup counts as neither a success nor a failure.
+func TestFetchAndUpdateMarksAnASWithoutReferral(t *testing.T) {
+	t.Parallel()
+
+	f, db := newFetcherWithWHOISAnswers(t,
+		map[string]string{ianaWHOISServer: ianaPrivateUseBlock})
+
+	f.fetchAndUpdate(t.Context(), testASNA)
+
+	asn, err := db.GetNextStaleASN(t.Context(), whoisStaleThreshold)
+	if !errors.Is(err, database.ErrNoStaleASN) {
+		t.Errorf("GetNextStaleASN = %d, %v; want ErrNoStaleASN", asn, err)
+	}
+
+	got := f.GetStats()
+	if got.CurrentInterval != baseInterval || got.SuccessesLastHour != 0 ||
+		got.ErrorsLastHour != 0 {
+		t.Errorf("interval %v, %d successes, %d errors; want %v, 0, 0",
+			got.CurrentInterval, got.SuccessesLastHour, got.ErrorsLastHour,
+			baseInterval)
+	}
+}
+
+// TestASNFetcherStopReturnsByItsDeadline starts a fetcher, has it look up an
+// AS whose WHOIS server does not answer until the test ends, and stops it with
+// a short deadline while that lookup is in progress. Stop must return all the
+// same instead of waiting for the lookup.
+func TestASNFetcherStopReturnsByItsDeadline(t *testing.T) {
+	t.Parallel()
+
+	const stopDeadline = 100 * time.Millisecond
+
+	lookupStarted := make(chan struct{})
+	release := make(chan struct{})
+
+	f := NewASNFetcher(newMockStore(), logger.New().Logger)
+	f.whoisClient.DialContext = func(
+		context.Context, string, string,
+	) (net.Conn, error) {
+		close(lookupStarted)
+		<-release
+
+		return nil, syscall.ECONNREFUSED
+	}
+
+	f.Start(t.Context())
+	f.QueueImmediate(testASNA)
+
+	select {
+	case <-lookupStarted:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the fetcher did not start the lookup")
+	}
+
+	ctx, cancel := context.WithTimeout(t.Context(), stopDeadline)
+	defer cancel()
+
+	if !returnsInTime(func() { f.Stop(ctx) }) {
+		t.Fatalf("Stop with a deadline of %s did not return while a lookup "+
+			"was in progress", stopDeadline)
+	}
+
+	close(release)
+
+	select {
+	case <-f.done:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the fetcher goroutine did not return after the lookup ended")
+	}
+}
+
+// TestRecordKeepsOnlyTheStatsWindow records a success and a failure every
+// second, by a clock the test sets, far longer than statsWindow, and never
+// asks for the stats. The fetcher must keep only the times within the last
+// statsWindow.
+func TestRecordKeepsOnlyTheStatsWindow(t *testing.T) {
+	t.Parallel()
+
+	const (
+		records = 100_000
+		step    = time.Second
+		want    = int(statsWindow / step)
+	)
+
+	clock := time.Now()
+
+	f := NewASNFetcher(newMockStore(), logger.New().Logger)
+	f.now = func() time.Time { return clock }
+
+	for range records {
+		clock = clock.Add(step)
+
+		f.recordSuccess()
+		f.recordFailure()
+	}
+
+	if len(f.successTimes) != want || len(f.errorTimes) != want {
+		t.Errorf("%d success and %d error times kept, want %d of each, "+
+			"those of the last %s", len(f.successTimes), len(f.errorTimes),
+			want, statsWindow)
+	}
+}

internal/routewatch/batchwrite_internal_test.go

@@ -0,0 +1,696 @@
+package routewatch
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"log/slog"
+	"slices"
+	"sync"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/ristypes"
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+const (
+	// handleMessageTimeout is how long a test waits for HandleMessage to
+	// return, or for a write to start, before it fails.
+	handleMessageTimeout = 2 * time.Second
+
+	// stillWaitingAfter is how long a test watches a HandleMessage that must
+	// wait. One that does not wait returns well within it.
+	stillWaitingAfter = 100 * time.Millisecond
+)
+
+// blockingStore is a mockStore whose batch writes for the AS, peer and prefix
+// handlers, delete of a peer's routes and peering write wait until the test
+// calls releaseWrites. writeStarted is closed when the first of them starts. It
+// records what each batch write was given.
+type blockingStore struct {
+	*mockStore
+
+	release      chan struct{}
+	releaseOnce  sync.Once
+	writeStarted chan struct{}
+	startOnce    sync.Once
+
+	mu       sync.Mutex
+	prefixes map[string]bool
+	asns     map[int]bool
+	peers    map[string]bool
+}
+
+func newBlockingStore() *blockingStore {
+	return &blockingStore{
+		mockStore:    newMockStore(),
+		release:      make(chan struct{}),
+		writeStarted: make(chan struct{}),
+		prefixes:     make(map[string]bool),
+		asns:         make(map[int]bool),
+		peers:        make(map[string]bool),
+	}
+}
+
+func (s *blockingStore) UpsertLiveRouteBatch(routes []*database.LiveRoute) error {
+	s.waitForRelease()
+	s.mu.Lock()
+	defer s.mu.Unlock()
+
+	for _, route := range routes {
+		s.prefixes[route.Prefix] = true
+	}
+
+	return nil
+}
+
+func (s *blockingStore) GetOrCreateASNBatch(asns map[int]database.SeenTimes) error {
+	s.waitForRelease()
+	s.mu.Lock()
+	defer s.mu.Unlock()
+
+	for asn := range asns {
+		s.asns[asn] = true
+	}
+
+	return nil
+}
+
+func (s *blockingStore) UpdatePeerBatch(peers map[string]database.PeerUpdate) error {
+	s.waitForRelease()
+	s.mu.Lock()
+	defer s.mu.Unlock()
+
+	for peerIP := range peers {
+		s.peers[peerIP] = true
+	}
+
+	return nil
+}
+
+func (s *blockingStore) DeleteLiveRoutesOfPeer(peerIP string) (int, error) {
+	s.waitForRelease()
+
+	return s.mockStore.DeleteLiveRoutesOfPeer(peerIP)
+}
+
+func (s *blockingStore) RecordPeering(
+	asA, asB int, seen database.SeenTimes,
+) error {
+	s.waitForRelease()
+
+	return s.mockStore.RecordPeering(asA, asB, seen)
+}
+
+func (s *blockingStore) waitForRelease() {
+	s.startOnce.Do(func() { close(s.writeStarted) })
+	<-s.release
+}
+
+// releaseWrites lets the held writes, and all later ones, finish. A second call
+// does nothing.
+func (s *blockingStore) releaseWrites() {
+	s.releaseOnce.Do(func() { close(s.release) })
+}
+
+// stopLog holds what a handler's Stop logs it did not write.
+//
+//nolint:tagliatelle // field names are the keys the handlers log
+type stopLog struct {
+	Msg          string `json:"msg"`
+	Updates      int    `json:"updates"`
+	PeersDown    int    `json:"peers_down"`
+	TableChanges int    `json:"table_changes"`
+	Paths        int    `json:"paths"`
+}
+
+// newErrorLogger returns a logger that writes only errors, as JSON lines, to
+// logs.
+func newErrorLogger(logs *bytes.Buffer) *logger.Logger {
+	return &logger.Logger{Logger: slog.New(slog.NewJSONHandler(logs,
+		&slog.HandlerOptions{Level: slog.LevelError}))}
+}
+
+// lastLogged decodes the last line in logs, which is the one Stop logs: a
+// handler logs nothing after it.
+func lastLogged(t *testing.T, logs *bytes.Buffer) stopLog {
+	t.Helper()
+
+	lines := bytes.Split(bytes.TrimSpace(logs.Bytes()), []byte("\n"))
+
+	var logged stopLog
+
+	err := json.Unmarshal(lines[len(lines)-1], &logged)
+	if err != nil {
+		t.Fatalf("want the error Stop logs, got %q: %v", logs.String(), err)
+	}
+
+	return logged
+}
+
+// returnsInTime reports whether f returns within handleMessageTimeout.
+func returnsInTime(f func()) bool {
+	done := make(chan struct{})
+
+	go func() {
+		f()
+		close(done)
+	}()
+
+	select {
+	case <-done:
+		return true
+	case <-time.After(handleMessageTimeout):
+		return false
+	}
+}
+
+// startBlockedWrite hands h the messages in fill, which fill its batch so that
+// the handler writes it to store, and waits until that write starts. store
+// holds it until the test calls releaseWrites.
+func startBlockedWrite(
+	t *testing.T,
+	store *blockingStore,
+	h interface {
+		HandleMessage(msg *ristypes.RISMessage)
+	},
+	fill []*ristypes.RISMessage,
+) {
+	t.Helper()
+
+	if !returnsInTime(func() {
+		for _, msg := range fill {
+			h.HandleMessage(msg)
+		}
+	}) {
+		t.Fatal("HandleMessage did not return after filling the batch: " +
+			"it waits for the database write")
+	}
+
+	select {
+	case <-store.writeStarted:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the handler did not write its full batch")
+	}
+}
+
+// handleWhileAWriteIsBlocked starts a blocked write with the messages in fill
+// and hands h the message extra while store holds the write. HandleMessage must
+// return each time. Then it lets the writes finish and stops h, which writes
+// what it holds.
+func handleWhileAWriteIsBlocked(
+	t *testing.T,
+	store *blockingStore,
+	h interface {
+		HandleMessage(msg *ristypes.RISMessage)
+		Stop(ctx context.Context)
+	},
+	fill []*ristypes.RISMessage,
+	extra *ristypes.RISMessage,
+) {
+	t.Helper()
+
+	defer h.Stop(t.Context())
+	defer store.releaseWrites()
+
+	startBlockedWrite(t, store, h, fill)
+
+	if !returnsInTime(func() { h.HandleMessage(extra) }) {
+		t.Fatal("HandleMessage did not return while a database write was blocked")
+	}
+}
+
+// handleAtMaxBatchSize starts a blocked write with the messages in fill, then,
+// while store holds the write, hands h the messages in toMax, which bring its
+// next batch to its maximum size, and the message extra. HandleMessage for
+// extra must wait until the test lets the write finish, and return after. Then
+// it stops h, which writes what it holds.
+func handleAtMaxBatchSize(
+	t *testing.T,
+	store *blockingStore,
+	h interface {
+		HandleMessage(msg *ristypes.RISMessage)
+		Stop(ctx context.Context)
+	},
+	fill, toMax []*ristypes.RISMessage,
+	extra *ristypes.RISMessage,
+) {
+	t.Helper()
+
+	defer h.Stop(t.Context())
+	defer store.releaseWrites()
+
+	startBlockedWrite(t, store, h, fill)
+
+	for _, msg := range toMax {
+		h.HandleMessage(msg)
+	}
+
+	returned := make(chan struct{})
+
+	go func() {
+		h.HandleMessage(extra)
+		close(returned)
+	}()
+
+	select {
+	case <-returned:
+		t.Fatal("HandleMessage returned while the batch was at its maximum " +
+			"size and a write was blocked")
+	case <-time.After(stillWaitingAfter):
+	}
+
+	store.releaseWrites()
+
+	select {
+	case <-returned:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("HandleMessage did not return after the database write finished")
+	}
+}
+
+// stopWhileAWriteIsBlocked starts a blocked write with the messages in fill,
+// hands h the message extra, and stops h with a deadline that passes while
+// store holds the write. Stop must return all the same. Then it lets the write
+// finish and waits until done, closed when h's flush goroutine returns. Past
+// the deadline that goroutine skips its last write, so extra is not written,
+// and Stop logs as not written both extra and what the held write took.
+func stopWhileAWriteIsBlocked(
+	t *testing.T,
+	store *blockingStore,
+	h interface {
+		HandleMessage(msg *ristypes.RISMessage)
+		Stop(ctx context.Context)
+	},
+	done <-chan struct{},
+	fill []*ristypes.RISMessage,
+	extra *ristypes.RISMessage,
+) {
+	t.Helper()
+
+	const stopDeadline = 100 * time.Millisecond
+
+	defer store.releaseWrites()
+
+	startBlockedWrite(t, store, h, fill)
+	h.HandleMessage(extra)
+
+	ctx, cancel := context.WithTimeout(t.Context(), stopDeadline)
+	defer cancel()
+
+	if !returnsInTime(func() { h.Stop(ctx) }) {
+		t.Fatalf("Stop with a deadline of %s did not return while a write "+
+			"was blocked", stopDeadline)
+	}
+
+	store.releaseWrites()
+
+	select {
+	case <-done:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the flush goroutine did not return after the write finished")
+	}
+}
+
+// TestPrefixHandlerSkipsItsLastWriteAfterTheStopDeadline also checks the
+// routing table's changes: the flush whose delete is held wrote those of the
+// fill, so Stop must log the one of the extra prefix as not written.
+func TestPrefixHandlerSkipsItsLastWriteAfterTheStopDeadline(t *testing.T) {
+	t.Parallel()
+
+	const extraPrefix = "198.51.100.0/24"
+
+	var logs bytes.Buffer
+
+	store := newBlockingStore()
+	h := NewPrefixHandler(store, newErrorLogger(&logs),
+		routetable.New(routetable.Params{}))
+	fill := slices.Repeat([]string{testV4Prefix}, prefixBatchSize)
+	// The peer's session goes down first, so the write held is the delete of
+	// its routes, and Stop must log that peer down too.
+	stopWhileAWriteIsBlocked(t, store, h, h.done,
+		[]*ristypes.RISMessage{sessionDown(), prefixAnnouncement(fill)},
+		prefixAnnouncement([]string{extraPrefix}))
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if store.prefixes[extraPrefix] {
+		t.Errorf("%s was written after the stop deadline", extraPrefix)
+	}
+
+	logged := lastLogged(t, &logs)
+	if logged.Updates != prefixBatchSize+1 || logged.PeersDown != 1 ||
+		logged.TableChanges != 1 {
+		t.Errorf("Stop logged %d updates, %d peer downs and %d table changes "+
+			"not written, want %d, 1 and 1", logged.Updates, logged.PeersDown,
+			logged.TableChanges, prefixBatchSize+1)
+	}
+}
+
+// tableWriteBlockingStore is a blockingStore whose writes of the routing
+// table's changes also wait until the test calls releaseWrites.
+type tableWriteBlockingStore struct {
+	*blockingStore
+}
+
+func (s tableWriteBlockingStore) WriteRouteChanges(
+	changes []routetable.Change, newPeers map[string]uint,
+) error {
+	s.waitForRelease()
+
+	return s.mockStore.WriteRouteChanges(changes, newPeers)
+}
+
+// TestPrefixHandlerStopCountsTheTableChangesOfAWriteInProgress holds the write
+// of the routing table's changes that a full batch starts, then hands the
+// prefix handler one more prefix and stops it with a deadline that passes
+// meanwhile. Stop must log as not written both the change the held write took
+// and the one since.
+func TestPrefixHandlerStopCountsTheTableChangesOfAWriteInProgress(t *testing.T) {
+	t.Parallel()
+
+	const extraPrefix = "198.51.100.0/24"
+
+	var logs bytes.Buffer
+
+	blocking := newBlockingStore()
+	h := NewPrefixHandler(tableWriteBlockingStore{blockingStore: blocking},
+		newErrorLogger(&logs), routetable.New(routetable.Params{}))
+	fill := slices.Repeat([]string{testV4Prefix}, prefixBatchSize)
+	stopWhileAWriteIsBlocked(t, blocking, h, h.done,
+		[]*ristypes.RISMessage{prefixAnnouncement(fill)},
+		prefixAnnouncement([]string{extraPrefix}))
+
+	if logged := lastLogged(t, &logs); logged.TableChanges != 2 {
+		t.Errorf("Stop logged %d table changes not written, want 2",
+			logged.TableChanges)
+	}
+}
+
+func TestASHandlerSkipsItsLastWriteAfterTheStopDeadline(t *testing.T) {
+	t.Parallel()
+
+	// Each message gives an update for its origin AS and one for each AS
+	// in its path: fill asnBatchSize+1, extra 2.
+	const wantUpdates = asnBatchSize + 1 + 2
+
+	var logs bytes.Buffer
+
+	store := newBlockingStore()
+	h := NewASHandler(store, newErrorLogger(&logs))
+	fill := ristypes.ASPath{Sequence: slices.Repeat([]int{testASNA}, asnBatchSize)}
+	stopWhileAWriteIsBlocked(t, store, h, h.done,
+		[]*ristypes.RISMessage{{Path: fill}},
+		&ristypes.RISMessage{Path: ristypes.ASPath{Sequence: []int{testASNB}}})
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if store.asns[testASNB] {
+		t.Errorf("AS%d was written after the stop deadline", testASNB)
+	}
+
+	if logged := lastLogged(t, &logs); logged.Updates != wantUpdates {
+		t.Errorf("Stop logged %d updates not written, want %d",
+			logged.Updates, wantUpdates)
+	}
+}
+
+func TestPeerHandlerSkipsItsLastWriteAfterTheStopDeadline(t *testing.T) {
+	t.Parallel()
+
+	const otherPeerIP = "2001:db8:ffff::2"
+
+	var logs bytes.Buffer
+
+	store := newBlockingStore()
+	h := NewPeerHandler(store, newErrorLogger(&logs))
+	stopWhileAWriteIsBlocked(t, store, h, h.done,
+		slices.Repeat([]*ristypes.RISMessage{{Peer: testPeerIP}}, peerBatchSize),
+		&ristypes.RISMessage{Peer: otherPeerIP})
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if store.peers[otherPeerIP] {
+		t.Errorf("peer %s was written after the stop deadline", otherPeerIP)
+	}
+
+	if logged := lastLogged(t, &logs); logged.Updates != peerBatchSize+1 {
+		t.Errorf("Stop logged %d updates not written, want %d",
+			logged.Updates, peerBatchSize+1)
+	}
+}
+
+// TestPeeringHandlerStopCountsTheWriteInProgress hands a running peering
+// handler one AS path and stops it with a deadline that passes while store
+// holds Stop's last write, of that path's peering. Stop must log that path as
+// not written.
+func TestPeeringHandlerStopCountsTheWriteInProgress(t *testing.T) {
+	t.Parallel()
+
+	const stopDeadline = 100 * time.Millisecond
+
+	var logs bytes.Buffer
+
+	store := newBlockingStore()
+	defer store.releaseWrites()
+
+	h := NewPeeringHandler(store, newErrorLogger(&logs))
+	handle(h, time.Now().UTC(), testASNA, testASNB)
+
+	ctx, cancel := context.WithTimeout(t.Context(), stopDeadline)
+	defer cancel()
+
+	if !returnsInTime(func() { h.Stop(ctx) }) {
+		t.Fatalf("Stop with a deadline of %s did not return while a write "+
+			"was blocked", stopDeadline)
+	}
+
+	store.releaseWrites()
+
+	select {
+	case <-h.done:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the processing goroutine did not return after the write " +
+			"finished")
+	}
+
+	if logged := lastLogged(t, &logs); logged.Paths != 1 {
+		t.Errorf("Stop logged %d paths not written, want 1", logged.Paths)
+	}
+}
+
+// prefixAnnouncement returns a message in which testPeerIP announces prefixes
+// with origin AS testASNA.
+func prefixAnnouncement(prefixes []string) *ristypes.RISMessage {
+	return &ristypes.RISMessage{
+		Peer: testPeerIP,
+		Path: ristypes.ASPath{Sequence: []int{testASNA}},
+		Announcements: []ristypes.RISAnnouncement{
+			{NextHop: testPeerIP, Prefixes: prefixes},
+		},
+	}
+}
+
+func TestPrefixHandlerHandlesMessagesWhileAWriteIsBlocked(t *testing.T) {
+	t.Parallel()
+
+	const extraPrefix = "198.51.100.0/24"
+
+	store := newBlockingStore()
+	h := NewPrefixHandler(store, logger.New(), routetable.New(routetable.Params{}))
+	fill := slices.Repeat([]string{testV4Prefix}, prefixBatchSize)
+	handleWhileAWriteIsBlocked(t, store, h,
+		[]*ristypes.RISMessage{prefixAnnouncement(fill)},
+		prefixAnnouncement([]string{extraPrefix}))
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if !store.prefixes[extraPrefix] {
+		t.Errorf("Stop did not write %s, handled while a write was blocked",
+			extraPrefix)
+	}
+}
+
+func TestASHandlerHandlesMessagesWhileAWriteIsBlocked(t *testing.T) {
+	t.Parallel()
+
+	store := newBlockingStore()
+	h := NewASHandler(store, logger.New())
+	fill := ristypes.ASPath{Sequence: slices.Repeat([]int{testASNA}, asnBatchSize)}
+	handleWhileAWriteIsBlocked(t, store, h,
+		[]*ristypes.RISMessage{{Path: fill}},
+		&ristypes.RISMessage{Path: ristypes.ASPath{Sequence: []int{testASNB}}})
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if !store.asns[testASNB] {
+		t.Errorf("Stop did not write AS%d, handled while a write was blocked",
+			testASNB)
+	}
+}
+
+func TestPeerHandlerHandlesMessagesWhileAWriteIsBlocked(t *testing.T) {
+	t.Parallel()
+
+	const otherPeerIP = "2001:db8:ffff::2"
+
+	store := newBlockingStore()
+	h := NewPeerHandler(store, logger.New())
+	handleWhileAWriteIsBlocked(t, store, h,
+		slices.Repeat([]*ristypes.RISMessage{{Peer: testPeerIP}}, peerBatchSize),
+		&ristypes.RISMessage{Peer: otherPeerIP})
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if !store.peers[otherPeerIP] {
+		t.Errorf("Stop did not write peer %s, handled while a write was blocked",
+			otherPeerIP)
+	}
+}
+
+func TestPrefixHandlerWaitsAtTheMaximumBatchSize(t *testing.T) {
+	t.Parallel()
+
+	const extraPrefix = "198.51.100.0/24"
+
+	store := newBlockingStore()
+	h := NewPrefixHandler(store, logger.New(), routetable.New(routetable.Params{}))
+	fill := slices.Repeat([]string{testV4Prefix}, prefixBatchSize)
+	toMax := slices.Repeat([]string{testV4Prefix}, prefixBatchMaxSize)
+	handleAtMaxBatchSize(t, store, h,
+		[]*ristypes.RISMessage{prefixAnnouncement(fill)},
+		[]*ristypes.RISMessage{prefixAnnouncement(toMax)},
+		prefixAnnouncement([]string{extraPrefix}))
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if !store.prefixes[extraPrefix] {
+		t.Errorf("Stop did not write %s, handled after the batch reached its "+
+			"maximum size", extraPrefix)
+	}
+}
+
+func TestASHandlerWaitsAtTheMaximumBatchSize(t *testing.T) {
+	t.Parallel()
+
+	store := newBlockingStore()
+	h := NewASHandler(store, logger.New())
+	fill := ristypes.ASPath{Sequence: slices.Repeat([]int{testASNA}, asnBatchSize)}
+	toMax := ristypes.ASPath{Sequence: slices.Repeat([]int{testASNA}, asnBatchMaxSize)}
+	handleAtMaxBatchSize(t, store, h,
+		[]*ristypes.RISMessage{{Path: fill}},
+		[]*ristypes.RISMessage{{Path: toMax}},
+		&ristypes.RISMessage{Path: ristypes.ASPath{Sequence: []int{testASNB}}})
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if !store.asns[testASNB] {
+		t.Errorf("Stop did not write AS%d, handled after the batch reached its "+
+			"maximum size", testASNB)
+	}
+}
+
+func TestPeerHandlerWaitsAtTheMaximumBatchSize(t *testing.T) {
+	t.Parallel()
+
+	const otherPeerIP = "2001:db8:ffff::2"
+
+	store := newBlockingStore()
+	h := NewPeerHandler(store, logger.New())
+	message := []*ristypes.RISMessage{{Peer: testPeerIP}}
+	handleAtMaxBatchSize(t, store, h,
+		slices.Repeat(message, peerBatchSize),
+		slices.Repeat(message, peerBatchMaxSize),
+		&ristypes.RISMessage{Peer: otherPeerIP})
+
+	store.mu.Lock()
+	defer store.mu.Unlock()
+
+	if !store.peers[otherPeerIP] {
+		t.Errorf("Stop did not write peer %s, handled after the batch reached "+
+			"its maximum size", otherPeerIP)
+	}
+}
+
+// TestHandlersWakeTheFlushGoroutineWhenTheBatchIsFull hands each handler, built
+// without its flush goroutine so that nothing takes a wake-up, one update short
+// of a full batch, and a second one exactly a full batch. Only the full batch
+// leaves a wake-up in batchFullCh.
+func TestHandlersWakeTheFlushGoroutineWhenTheBatchIsFull(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name      string
+		batchSize int
+		// handle hands a new handler n updates and returns its batchFullCh.
+		handle func(n int) chan struct{}
+	}{
+		{
+			name:      "prefix",
+			batchSize: prefixBatchSize,
+			handle: func(n int) chan struct{} {
+				h := &PrefixHandler{
+					batchFullCh: make(chan struct{}, 1),
+					table:       routetable.New(routetable.Params{}),
+				}
+				h.HandleMessage(prefixAnnouncement(
+					slices.Repeat([]string{testV4Prefix}, n)))
+
+				return h.batchFullCh
+			},
+		},
+		{
+			name:      "AS",
+			batchSize: asnBatchSize,
+			handle: func(n int) chan struct{} {
+				h := &ASHandler{batchFullCh: make(chan struct{}, 1)}
+				// The origin AS and each AS in the path are one update each.
+				h.HandleMessage(&ristypes.RISMessage{
+					Path: ristypes.ASPath{Sequence: slices.Repeat([]int{testASNA}, n-1)},
+				})
+
+				return h.batchFullCh
+			},
+		},
+		{
+			name:      "peer",
+			batchSize: peerBatchSize,
+			handle: func(n int) chan struct{} {
+				h := &PeerHandler{batchFullCh: make(chan struct{}, 1)}
+				for range n {
+					h.HandleMessage(&ristypes.RISMessage{Peer: testPeerIP})
+				}
+
+				return h.batchFullCh
+			},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			got := len(tt.handle(tt.batchSize - 1))
+			if got != 0 {
+				t.Errorf("one update short of a full batch left %d wake-ups, "+
+					"want none", got)
+			}
+
+			got = len(tt.handle(tt.batchSize))
+			if got != 1 {
+				t.Errorf("a full batch left %d wake-ups, want 1", got)
+			}
+		})
+	}
+}

internal/routewatch/cli.go

@@ -3,10 +3,8 @@ package routewatch
 import (
 	"context"
 	"os"
-	"os/signal"
 	"runtime"
 	"strings"
-	"syscall"
 	"time"
 
 	"git.eeqj.de/sneak/routewatch/internal/logger"
@@ -14,12 +12,20 @@ import (
 )
 
 const (
-	// shutdownTimeout is the maximum time allowed for graceful shutdown
+	// shutdownTimeout is how long Shutdown has after a stop signal. upaas,
-	shutdownTimeout = 60 * time.Second
+	// like docker stop by default, kills the container 10 seconds after the
+	// stop signal (README "Running under upaas").
+	shutdownTimeout = 8 * time.Second
+	// stopTimeout is fx's limit on the stop hook, past which fx exits with
+	// status 1: shutdownTimeout and time for Shutdown to close the HTTP
+	// server after its deadline.
+	stopTimeout = 9 * time.Second
 	// debugInterval is how often to log debug stats
 	debugInterval = 60 * time.Second
 	// bytesPerMB is bytes per megabyte
 	bytesPerMB = 1024 * 1024
+	// failureExitCode is the exit status when RouteWatch stops on an error
+	failureExitCode = 1
 )
 
 // logDebugStats logs goroutine count and memory usage
@@ -54,43 +60,47 @@ func logDebugStats(logger *logger.Logger) {
 }
 
 // CLIEntry is the main entry point for the routewatch command-line interface.
-// It initializes the application using the fx dependency injection framework,
+// It initializes the application using the fx dependency injection framework
-// sets up signal handling for graceful shutdown, and starts the RouteWatch service.
+// and starts the RouteWatch service. It blocks until fx receives SIGINT or
-// This function blocks until the application receives a shutdown signal or encounters
+// SIGTERM or RouteWatch stops on an error; fx then runs the stop hook and sets
-// a fatal error.
+// the exit code.
 func CLIEntry() {
 	app := fx.New(
-		getModule(),
+		fx.Options(getModule()...),
-		fx.StopTimeout(shutdownTimeout), // Allow 60 seconds for graceful shutdown
+		fx.StopTimeout(stopTimeout),
-		fx.Invoke(func(lc fx.Lifecycle, rw *RouteWatch, logger *logger.Logger, shutdowner fx.Shutdowner) {
+		fx.Invoke(func(
+			lc fx.Lifecycle,
+			rw *RouteWatch,
+			logger *logger.Logger,
+			shutdowner fx.Shutdowner,
+		) {
 			lc.Append(fx.Hook{
 				OnStart: func(ctx context.Context) error {
 					// Start debug stats logging
 					go logDebugStats(logger)
 
-					// Handle shutdown signals
-					sigCh := make(chan os.Signal, 1)
-					signal.Notify(sigCh, os.Interrupt, syscall.SIGTERM)
-
 					go func() {
-						<-sigCh
+						err := rw.Run(ctx)
-						logger.Info("Received shutdown signal")
+						if err != nil {
-						if err := shutdowner.Shutdown(); err != nil {
-							logger.Error("Failed to shutdown gracefully", "error", err)
-						}
-					}()
-
-					go func() {
-						if err := rw.Run(ctx); err != nil {
 							logger.Error("RouteWatch error", "error", err)
+
+							err = shutdowner.Shutdown(fx.ExitCode(failureExitCode))
+							if err != nil {
+								logger.Error("Failed to shutdown gracefully",
+									"error", err)
+							}
 						}
 					}()
 
 					return nil
 				},
-				OnStop: func(_ context.Context) error {
+				OnStop: func(ctx context.Context) error {
 					logger.Info("Shutting down RouteWatch")
-					rw.Shutdown()
+
+					ctx, cancel := context.WithTimeout(ctx, shutdownTimeout)
+					defer cancel()
+
+					rw.Shutdown(ctx)
 
 					return nil
 				},

internal/routewatch/dbmaintainer.go

@@ -1,8 +1,8 @@
-// Package routewatch contains the database maintainer for background maintenance tasks.
 package routewatch
 
 import (
 	"context"
+	"errors"
 	"log/slog"
 	"sync"
 	"time"
@@ -35,12 +35,18 @@ const (
 	analyzeTimeout = 5 * time.Minute
 )
 
+// checkpointIncompleteMessage is logged, at debug level like a completed
+// checkpoint, when a checkpoint leaves part of the WAL for a later one: while
+// reads overlap, most do. The database warns when the WAL file stays over its
+// size limit.
+const checkpointIncompleteMessage = "WAL checkpoint incomplete"
+
 // DBMaintainer handles background database maintenance tasks.
 type DBMaintainer struct {
 	db     database.Store
 	logger *slog.Logger
 	stopCh chan struct{}
-	wg     sync.WaitGroup
+	done   chan struct{} // closed when the maintenance goroutine returns
 
 	// Stats tracking
 	statsMu             sync.Mutex
@@ -61,13 +67,15 @@ func NewDBMaintainer(db database.Store, logger *slog.Logger) *DBMaintainer {
 		db:     db,
 		logger: logger.With("component", "db_maintainer"),
 		stopCh: make(chan struct{}),
+		done:   make(chan struct{}),
 	}
 }
 
-// Start begins the background maintenance goroutine.
+// Start begins the background maintenance goroutine. It runs until Stop:
-func (m *DBMaintainer) Start() {
+// the end of ctx does not stop it.
-	m.wg.Add(1)
+func (m *DBMaintainer) Start(ctx context.Context) {
-	go m.run()
+	go m.run(context.WithoutCancel(ctx))
+
 	m.logger.Info("Database maintainer started",
 		"checkpoint_interval", checkpointInterval,
 		"vacuum_interval", vacuumInterval,
@@ -75,23 +83,32 @@ func (m *DBMaintainer) Start() {
 	)
 }
 
-// Stop gracefully shuts down the maintainer.
+// Stop stops the maintainer and waits, until ctx ends, for a maintenance task
-func (m *DBMaintainer) Stop() {
+// in progress to finish.
+func (m *DBMaintainer) Stop(ctx context.Context) {
 	close(m.stopCh)
-	m.wg.Wait()
+
+	select {
+	case <-m.done:
 		m.logger.Info("Database maintainer stopped")
+	case <-ctx.Done():
+		m.logger.Warn("Stop deadline passed before the database maintainer " +
+			"stopped")
+	}
 }
 
 // run is the main background loop.
-func (m *DBMaintainer) run() {
+func (m *DBMaintainer) run(ctx context.Context) {
-	defer m.wg.Done()
+	defer close(m.done)
 
 	// Use different timers for each task
 	checkpointTimer := time.NewTimer(checkpointInterval)
-	vacuumTimer := time.NewTimer(vacuumInterval)
-	analyzeTimer := time.NewTimer(analyzeInterval)
 	defer checkpointTimer.Stop()
+
+	vacuumTimer := time.NewTimer(vacuumInterval)
 	defer vacuumTimer.Stop()
+
+	analyzeTimer := time.NewTimer(analyzeInterval)
 	defer analyzeTimer.Stop()
 
 	for {
@@ -100,90 +117,92 @@ func (m *DBMaintainer) run() {
 			return
 
 		case <-checkpointTimer.C:
-			m.runCheckpoint()
+			m.runCheckpoint(ctx)
 			checkpointTimer.Reset(checkpointInterval)
 
 		case <-vacuumTimer.C:
-			m.runVacuum()
+			m.runVacuum(ctx)
 			vacuumTimer.Reset(vacuumInterval)
 
 		case <-analyzeTimer.C:
-			m.runAnalyze()
+			m.runAnalyze(ctx)
 			analyzeTimer.Reset(analyzeInterval)
 		}
 	}
 }
 
 // runCheckpoint performs a WAL checkpoint to keep the WAL file small.
-func (m *DBMaintainer) runCheckpoint() {
+func (m *DBMaintainer) runCheckpoint(ctx context.Context) {
-	ctx, cancel := context.WithTimeout(context.Background(), checkpointTimeout)
+	duration, err := runWithTimeout(ctx, checkpointTimeout, m.db.Checkpoint)
-	defer cancel()
+	m.recordResult(&m.lastCheckpoint, &m.lastCheckpointError, &m.checkpointCount, err)
 
-	startTime := time.Now()
+	switch {
-
+	case errors.Is(err, database.ErrCheckpointIncomplete):
-	err := m.db.Checkpoint(ctx)
+		m.logger.Debug(checkpointIncompleteMessage,
-
+			"error", err, "duration", duration)
-	m.statsMu.Lock()
+	case err != nil:
-	m.lastCheckpoint = time.Now()
+		m.logger.Error("WAL checkpoint failed", "error", err, "duration", duration)
-	m.lastCheckpointError = err
+	default:
-	if err == nil {
+		m.logger.Debug("WAL checkpoint completed", "duration", duration)
-		m.checkpointCount++
-	}
-	m.statsMu.Unlock()
-
-	if err != nil {
-		m.logger.Error("WAL checkpoint failed", "error", err, "duration", time.Since(startTime))
-	} else {
-		m.logger.Debug("WAL checkpoint completed", "duration", time.Since(startTime))
 	}
 }
 
 // runVacuum performs an incremental vacuum operation on the database.
-func (m *DBMaintainer) runVacuum() {
+func (m *DBMaintainer) runVacuum(ctx context.Context) {
-	ctx, cancel := context.WithTimeout(context.Background(), vacuumTimeout)
-	defer cancel()
-
 	m.logger.Debug("Running incremental vacuum")
-	startTime := time.Now()
 
-	err := m.db.Vacuum(ctx)
+	duration, err := runWithTimeout(ctx, vacuumTimeout, m.db.Vacuum)
-
+	m.recordResult(&m.lastVacuum, &m.lastVacuumError, &m.vacuumCount, err)
-	m.statsMu.Lock()
-	m.lastVacuum = time.Now()
-	m.lastVacuumError = err
-	if err == nil {
-		m.vacuumCount++
-	}
-	m.statsMu.Unlock()
 
 	if err != nil {
-		m.logger.Error("Incremental vacuum failed", "error", err, "duration", time.Since(startTime))
+		m.logger.Error("Incremental vacuum failed", "error", err, "duration", duration)
 	} else {
-		m.logger.Debug("Incremental vacuum completed", "duration", time.Since(startTime))
+		m.logger.Debug("Incremental vacuum completed", "duration", duration)
 	}
 }
 
 // runAnalyze performs an ANALYZE operation on the database.
-func (m *DBMaintainer) runAnalyze() {
+func (m *DBMaintainer) runAnalyze(ctx context.Context) {
-	ctx, cancel := context.WithTimeout(context.Background(), analyzeTimeout)
-	defer cancel()
-
 	m.logger.Info("Starting database ANALYZE")
-	startTime := time.Now()
 
-	err := m.db.Analyze(ctx)
+	duration, err := runWithTimeout(ctx, analyzeTimeout, m.db.Analyze)
-
+	m.recordResult(&m.lastAnalyze, &m.lastAnalyzeError, &m.analyzeCount, err)
-	m.statsMu.Lock()
-	m.lastAnalyze = time.Now()
-	m.lastAnalyzeError = err
-	if err == nil {
-		m.analyzeCount++
-	}
-	m.statsMu.Unlock()
 
 	if err != nil {
-		m.logger.Error("ANALYZE failed", "error", err, "duration", time.Since(startTime))
+		m.logger.Error("ANALYZE failed", "error", err, "duration", duration)
 	} else {
-		m.logger.Info("ANALYZE completed", "duration", time.Since(startTime))
+		m.logger.Info("ANALYZE completed", "duration", duration)
+	}
+}
+
+// runWithTimeout runs operation with a context that ends after timeout and
+// returns how long it took and its error.
+func runWithTimeout(
+	ctx context.Context,
+	timeout time.Duration,
+	operation func(context.Context) error,
+) (time.Duration, error) {
+	ctx, cancel := context.WithTimeout(ctx, timeout)
+	defer cancel()
+
+	startTime := time.Now()
+	err := operation(ctx)
+
+	return time.Since(startTime), err
+}
+
+// recordResult stores when a maintenance task last ran and its error, and
+// counts it when it succeeded.
+func (m *DBMaintainer) recordResult(
+	lastRun *time.Time, lastErr *error, successCount *int, err error,
+) {
+	m.statsMu.Lock()
+	defer m.statsMu.Unlock()
+
+	*lastRun = time.Now()
+	*lastErr = err
+
+	if err == nil {
+		*successCount++
 	}
 }

internal/routewatch/dbmaintainer_internal_test.go

@@ -0,0 +1,120 @@
+package routewatch
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"log/slog"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+)
+
+// incompleteCheckpointStore is a store whose every checkpoint leaves part of
+// the WAL uncopied.
+type incompleteCheckpointStore struct{ *mockStore }
+
+// Checkpoint reports the checkpoint incomplete.
+func (incompleteCheckpointStore) Checkpoint(context.Context) error {
+	return database.ErrCheckpointIncomplete
+}
+
+// TestIncompleteCheckpointNotLoggedAsCompleted runs a checkpoint that leaves
+// part of the WAL uncopied and checks that the maintainer logs it as
+// incomplete, not as completed, and does not count it as a checkpoint that
+// completed.
+func TestIncompleteCheckpointNotLoggedAsCompleted(t *testing.T) {
+	t.Parallel()
+
+	var logs bytes.Buffer
+
+	logger := slog.New(slog.NewJSONHandler(&logs,
+		&slog.HandlerOptions{Level: slog.LevelDebug}))
+	m := NewDBMaintainer(incompleteCheckpointStore{newMockStore()}, logger)
+
+	m.runCheckpoint(t.Context())
+
+	var logged struct {
+		Msg string `json:"msg"`
+	}
+
+	err := json.Unmarshal(logs.Bytes(), &logged)
+	if err != nil {
+		t.Fatalf("want one log message, got %q: %v", logs.String(), err)
+	}
+
+	if logged.Msg != checkpointIncompleteMessage {
+		t.Errorf("logged %q, want %q", logged.Msg, checkpointIncompleteMessage)
+	}
+
+	if m.checkpointCount != 0 {
+		t.Errorf("checkpointCount = %d after an incomplete checkpoint, want 0",
+			m.checkpointCount)
+	}
+}
+
+// heldCheckpointStore is a store whose checkpoint closes started and then
+// waits until release is closed.
+type heldCheckpointStore struct {
+	*mockStore
+
+	started chan struct{}
+	release chan struct{}
+}
+
+// Checkpoint waits until release is closed.
+func (s heldCheckpointStore) Checkpoint(context.Context) error {
+	close(s.started)
+	<-s.release
+
+	return nil
+}
+
+// TestDBMaintainerStopReturnsByItsDeadline stops the maintainer with a short
+// deadline while its goroutine runs a checkpoint that the store holds until
+// the test ends. Stop must return all the same instead of waiting for the
+// checkpoint.
+func TestDBMaintainerStopReturnsByItsDeadline(t *testing.T) {
+	t.Parallel()
+
+	const stopDeadline = 100 * time.Millisecond
+
+	store := heldCheckpointStore{
+		mockStore: newMockStore(),
+		started:   make(chan struct{}),
+		release:   make(chan struct{}),
+	}
+	m := NewDBMaintainer(store, slog.Default())
+
+	// Start would run the first checkpoint only after checkpointInterval.
+	// This goroutine stands for the one Start runs and checkpoints at once.
+	go func() {
+		defer close(m.done)
+
+		m.runCheckpoint(t.Context())
+	}()
+
+	select {
+	case <-store.started:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the maintainer did not start the checkpoint")
+	}
+
+	ctx, cancel := context.WithTimeout(t.Context(), stopDeadline)
+	defer cancel()
+
+	if !returnsInTime(func() { m.Stop(ctx) }) {
+		t.Fatalf("Stop with a deadline of %s did not return while a "+
+			"checkpoint was in progress", stopDeadline)
+	}
+
+	close(store.release)
+
+	select {
+	case <-m.done:
+	case <-time.After(handleMessageTimeout):
+		t.Fatal("the maintenance goroutine did not return after the " +
+			"checkpoint ended")
+	}
+}

internal/routewatch/handler.go

@@ -1,6 +1,8 @@
 package routewatch
 
 import (
+	"slices"
+
 	"git.eeqj.de/sneak/routewatch/internal/logger"
 	"git.eeqj.de/sneak/routewatch/internal/ristypes"
 )
@@ -15,10 +17,11 @@ type SimpleHandler struct {
 	callback     func(*ristypes.RISMessage)
 }
 
-// NewSimpleHandler creates a new SimpleHandler that accepts specific message types.
+// NewSimpleHandler creates a new SimpleHandler that accepts specific message
-// The messageTypes parameter specifies which RIS message types this handler will process.
+// types. The messageTypes parameter specifies which RIS message types this
-// If messageTypes is empty, the handler will accept all message types.
+// handler will process. If messageTypes is empty, the handler will accept all
-// The callback function is invoked for each message that passes the type filter.
+// message types. The callback function is invoked for each message that
+// passes the type filter.
 func NewSimpleHandler(
 	logger *logger.Logger,
 	messageTypes []string,
@@ -31,22 +34,17 @@ func NewSimpleHandler(
 	}
 }
 
-// WantsMessage returns true if this handler wants to process messages of the given type.
+// WantsMessage returns true if this handler wants to process messages of the
-// It checks whether messageType is in the handler's configured list of accepted types.
+// given type. It checks whether messageType is in the handler's configured
-// If no specific types were configured (empty messageTypes slice), it returns true for all types.
+// list of accepted types. If no specific types were configured (empty
+// messageTypes slice), it returns true for all types.
 func (h *SimpleHandler) WantsMessage(messageType string) bool {
 	// If no specific types are set, accept all messages
 	if len(h.messageTypes) == 0 {
 		return true
 	}
 
-	for _, t := range h.messageTypes {
+	return slices.Contains(h.messageTypes, messageType)
-		if t == messageType {
-			return true
-		}
-	}
-
-	return false
 }
 
 // HandleMessage processes a RIS message by invoking the configured callback function.

internal/routewatch/peerhandler.go

@@ -1,9 +1,11 @@
 package routewatch
 
-// peerhandler.go provides batched peer tracking functionality for BGP route monitoring.
+// peerhandler.go provides batched peer tracking functionality for BGP route
-// It tracks BGP peers from all incoming RIS messages and maintains peer state in the database.
+// monitoring. It tracks BGP peers from incoming UPDATE messages and maintains
+// peer state in the database.
 
 import (
+	"context"
 	"strconv"
 	"sync"
 	"time"
@@ -23,24 +25,42 @@ const (
 	// peerBatchSize is the number of peer updates to batch together
 	peerBatchSize = 10000
 
+	// peerBatchMaxSize is the most peer updates the batch holds while the
+	// flush goroutine writes the previous one: ten batches, about 10 MiB
+	// (README "Memory"). At this size HandleMessage waits until the flush
+	// goroutine takes the batch.
+	peerBatchMaxSize = 100000
+
 	// peerBatchTimeout is the maximum time to wait before flushing a batch
 	peerBatchTimeout = 2 * time.Second
 )
 
-// PeerHandler tracks BGP peers from all message types using batched operations.
+// PeerHandler tracks BGP peers from UPDATE messages using batched operations.
 // It maintains a queue of peer updates and periodically flushes them to the database
-// in batches to improve performance. The handler deduplicates peer updates within
+// in batches to improve performance. The handler writes one row per peer IP
-// each batch, keeping only the most recent update for each peer IP address.
+// address for each batch, with the earliest and the latest time of its updates.
 type PeerHandler struct {
 	db     database.Store
 	logger *logger.Logger
 
-	// Batching
+	// Batching. mu guards peerBatch and writing only: the flush goroutine
+	// takes the batch under mu and writes it after releasing mu, so
+	// HandleMessage waits for a database write only while peerBatch is at
+	// peerBatchMaxSize.
 	mu        sync.Mutex
 	peerBatch []peerUpdate
-	lastFlush time.Time
+	// writing is how many updates the flush goroutine took from peerBatch
-	stopCh    chan struct{}
+	// and has not finished writing.
-	wg        sync.WaitGroup
+	writing int
+	// batchTaken is signalled when the flush goroutine takes peerBatch; its L
+	// is &mu.
+	batchTaken  sync.Cond
+	batchFullCh chan struct{} // wakes the flush goroutine when peerBatch is full
+	lastFlush   time.Time     // used only by the flush goroutine
+	// stopCh hands Stop's context to the flush goroutine. It holds one, so
+	// Stop does not wait for a write in progress to hand it over.
+	stopCh chan context.Context
+	done   chan struct{} // closed when the flush goroutine returns
 }
 
 type peerUpdate struct {
@@ -59,22 +79,24 @@ func NewPeerHandler(db database.Store, logger *logger.Logger) *PeerHandler {
 		db:          db,
 		logger:      logger,
 		peerBatch:   make([]peerUpdate, 0, peerBatchSize),
+		batchFullCh: make(chan struct{}, 1),
 		lastFlush:   time.Now(),
-		stopCh:    make(chan struct{}),
+		stopCh:      make(chan context.Context, 1),
+		done:        make(chan struct{}),
 	}
+	h.batchTaken.L = &h.mu
 
 	// Start the flush timer goroutine
-	h.wg.Add(1)
 	go h.flushLoop()
 
 	return h
 }
 
-// WantsMessage returns true for all message types since peer information
+// WantsMessage returns true for UPDATE messages only. A peer whose session
-// is extracted from every RIS message regardless of type. This satisfies
+// never comes up sends STATE messages and nothing else, so a STATE message is
-// the MessageHandler interface.
+// not a sign of a peer that carries routes.
-func (h *PeerHandler) WantsMessage(_ string) bool {
+func (h *PeerHandler) WantsMessage(messageType string) bool {
-	return true
+	return messageType == ristypes.MessageTypeUpdate
 }
 
 // QueueCapacity returns the desired queue capacity for this handler.
@@ -87,13 +109,15 @@ func (h *PeerHandler) QueueCapacity() int {
 
 // HandleMessage processes a RIS message to track peer information.
 // It extracts the peer IP address and ASN from the message and adds
-// the update to an internal batch. When the batch reaches peerBatchSize
+// the update to an internal batch. The flush goroutine writes the batch to the
-// or the batch timeout expires, the batch is flushed to the database.
+// database when it reaches peerBatchSize or the batch timeout expires.
 func (h *PeerHandler) HandleMessage(msg *ristypes.RISMessage) {
 	// Parse peer ASN from string
 	peerASN := 0
+
 	if msg.PeerASN != "" {
-		if asn, err := strconv.Atoi(msg.PeerASN); err == nil {
+		asn, err := strconv.Atoi(msg.PeerASN)
+		if err == nil {
 			peerASN = asn
 		}
 	}
@@ -101,6 +125,13 @@ func (h *PeerHandler) HandleMessage(msg *ristypes.RISMessage) {
 	h.mu.Lock()
 	defer h.mu.Unlock()
 
+	// While the batch is at its maximum size, wait until the flush goroutine
+	// takes it. Meanwhile the handler takes nothing off its queue, which
+	// fills, and the streamer drops messages for it.
+	for len(h.peerBatch) >= peerBatchMaxSize {
+		h.batchTaken.Wait()
+	}
+
 	// Add to batch
 	h.peerBatch = append(h.peerBatch, peerUpdate{
 		peerIP:      msg.Peer,
@@ -109,74 +140,120 @@ func (h *PeerHandler) HandleMessage(msg *ristypes.RISMessage) {
 		timestamp:   msg.ParsedTimestamp,
 	})
 
-	// Check if we need to flush
+	// A full batch wakes the flush goroutine. If it is writing, the wake-up
+	// waits in batchFullCh and the batch keeps growing until the write ends
+	// or the batch reaches peerBatchMaxSize.
 	if len(h.peerBatch) >= peerBatchSize {
-		h.flushBatchLocked()
+		select {
+		case h.batchFullCh <- struct{}{}:
+		default: // already woken
+		}
 	}
 }
 
-// flushLoop runs in a goroutine and periodically flushes batches
+// Stop stops the flush goroutine, which writes the updates the handler holds
+// before it returns, and waits for it until ctx ends. Once ctx has ended the
+// flush goroutine skips that last write, and a write still running is lost
+// when the process exits, so Stop logs how many updates are not written.
+func (h *PeerHandler) Stop(ctx context.Context) {
+	h.stopCh <- ctx
+
+	select {
+	case <-h.done:
+	case <-ctx.Done():
+	}
+
+	h.mu.Lock()
+	unwritten := h.writing + len(h.peerBatch)
+	h.mu.Unlock()
+
+	if unwritten > 0 {
+		h.logger.Error("Stop deadline passed before the peer handler wrote "+
+			"its updates", "updates", unwritten)
+	}
+}
+
+// flushLoop runs in a goroutine, the only one that writes the batch: it
+// flushes when the batch is full, periodically, and once more on Stop.
 func (h *PeerHandler) flushLoop() {
-	defer h.wg.Done()
+	defer close(h.done)
+
 	ticker := time.NewTicker(peerBatchTimeout)
 	defer ticker.Stop()
 
 	for {
 		select {
 		case <-ticker.C:
-			h.mu.Lock()
 			if time.Since(h.lastFlush) >= peerBatchTimeout {
-				h.flushBatchLocked()
+				h.flushBatch()
 			}
-			h.mu.Unlock()
+		case <-h.batchFullCh:
-		case <-h.stopCh:
+			h.flushBatch()
-			// Final flush
+		case ctx := <-h.stopCh:
+			// Final flush, unless the stop deadline has passed
+			if ctx.Err() == nil {
+				h.flushBatch()
+			}
+
+			return
+		}
+	}
+}
+
+// takeBatch returns the pending updates and leaves an empty batch in their
+// place for HandleMessage to fill. They count in writing until flushBatch has
+// written them.
+func (h *PeerHandler) takeBatch() []peerUpdate {
 	h.mu.Lock()
-			h.flushBatchLocked()
+	defer h.mu.Unlock()
-			h.mu.Unlock()
 
-			return
-		}
-	}
-}
-
-// flushBatchLocked flushes the peer batch to the database (must be called with mutex held)
-func (h *PeerHandler) flushBatchLocked() {
 	if len(h.peerBatch) == 0 {
+		return nil
+	}
+
+	batch := h.peerBatch
+	h.writing = len(batch)
+	h.peerBatch = make([]peerUpdate, 0, peerBatchSize)
+	h.batchTaken.Broadcast()
+
+	return batch
+}
+
+// flushBatch takes the pending batch and writes it to the database, without
+// holding the mutex during the write.
+func (h *PeerHandler) flushBatch() {
+	batch := h.takeBatch()
+	if len(batch) == 0 {
 		return
 	}
 
-	// Deduplicate by peer IP, keeping the latest update for each peer
+	// The earliest and the latest time the batch saw each peer, with the peer
-	peerMap := make(map[string]peerUpdate)
+	// AS and message type of its latest update. The batch holds several
-	for _, update := range h.peerBatch {
+	// updates for a peer, so a new peer is stored first seen at the earliest of
-		if existing, ok := peerMap[update.peerIP]; !ok || update.timestamp.After(existing.timestamp) {
+	// them and last seen at the latest.
-			peerMap[update.peerIP] = update
+	peers := make(map[string]database.PeerUpdate)
-		}
+	for _, update := range batch {
+		peer, ok := peers[update.peerIP]
+		if !ok || update.timestamp.After(peer.Seen.Last) {
+			peer.PeerIP = update.peerIP
+			peer.PeerASN = update.peerASN
+			peer.MessageType = update.messageType
 		}
 
-	// Convert to database format
+		peer.Seen.Add(update.timestamp)
-	dbPeerMap := make(map[string]database.PeerUpdate)
+		peers[update.peerIP] = peer
-	for peerIP, update := range peerMap {
-		dbPeerMap[peerIP] = database.PeerUpdate{
-			PeerIP:      update.peerIP,
-			PeerASN:     update.peerASN,
-			MessageType: update.messageType,
-			Timestamp:   update.timestamp,
-		}
 	}
 
 	// Process all peers in a single batch transaction
-	if err := h.db.UpdatePeerBatch(dbPeerMap); err != nil {
+	err := h.db.UpdatePeerBatch(peers)
-		h.logger.Error("Failed to process peer batch", "error", err, "count", len(dbPeerMap))
+	if err != nil {
+		h.logger.Error("Failed to process peer batch",
+			"error", err, "count", len(peers))
 	}
 
-	// Clear batch
+	h.mu.Lock()
-	h.peerBatch = h.peerBatch[:0]
+	h.writing = 0
+	h.mu.Unlock()
+
 	h.lastFlush = time.Now()
 }
-
-// Stop gracefully stops the handler and flushes remaining batches
-func (h *PeerHandler) Stop() {
-	close(h.stopCh)
-	h.wg.Wait()
-}

internal/routewatch/peerhandler_internal_test.go

@@ -0,0 +1,59 @@
+package routewatch
+
+import (
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+)
+
+// peerBatchStore keeps the peers of the last UpdatePeerBatch call. The
+// database has no read of the peer table, so the peer handler's test checks
+// what the handler writes.
+type peerBatchStore struct {
+	*mockStore
+
+	peers map[string]database.PeerUpdate
+}
+
+func (s *peerBatchStore) UpdatePeerBatch(
+	peers map[string]database.PeerUpdate,
+) error {
+	s.peers = peers
+
+	return nil
+}
+
+// TestPeerHandlerWritesTheEarliestAndLatestTimesOfANewPeer handles two updates
+// from one peer, the second 5 seconds after the first, in one batch. The peer
+// handler writes the peer first seen at the time of the first update and last
+// seen at the time of the second.
+func TestPeerHandlerWritesTheEarliestAndLatestTimesOfANewPeer(t *testing.T) {
+	t.Parallel()
+
+	store := &peerBatchStore{mockStore: newMockStore()}
+
+	// Built without NewPeerHandler's flush goroutine, so the test flushes
+	// itself.
+	h := &PeerHandler{db: store, logger: logger.New()}
+
+	first := peerAnnouncement(testPeerIP, testV4Prefix)
+	second := peerAnnouncement(testPeerIP, testV4Prefix)
+	second.ParsedTimestamp = first.ParsedTimestamp.Add(secondUpdateAfter)
+
+	h.HandleMessage(first)
+	h.HandleMessage(second)
+	h.flushBatch()
+
+	seen := store.peers[testPeerIP].Seen
+
+	if !seen.First.Equal(first.ParsedTimestamp) {
+		t.Errorf("first_seen of peer %s = %s, want %s", testPeerIP,
+			seen.First, first.ParsedTimestamp)
+	}
+
+	if !seen.Last.Equal(second.ParsedTimestamp) {
+		t.Errorf("last_seen of peer %s = %s, want %s", testPeerIP,
+			seen.Last, second.ParsedTimestamp)
+	}
+}

internal/routewatch/peeringhandler.go

@@ -1,6 +1,7 @@
 package routewatch
 
 import (
+	"context"
 	"encoding/json"
 	"sync"
 	"time"
@@ -41,12 +42,19 @@ type PeeringHandler struct {
 	db     database.Store
 	logger *logger.Logger
 
-	// In-memory AS path tracking
+	// In-memory AS path tracking: each path, JSON-encoded, with the earliest
+	// and the latest time an update carried it.
 	mu           sync.Mutex
-	asPaths      map[string]time.Time // key is JSON-encoded AS path
+	asPaths      map[string]database.SeenTimes
-	droppedPaths int                  // paths dropped because the map was full
+	droppedPaths int // paths dropped since the last run because the map was full
+	// writing is how many paths processPeerings took from asPaths and has
+	// not finished writing the peerings of.
+	writing int
 
-	stopCh chan struct{}
+	// stopCh hands Stop's context to the processing goroutine. It holds one,
+	// so Stop does not wait for a write in progress to hand it over.
+	stopCh chan context.Context
+	done   chan struct{} // closed when the processing goroutine returns
 }
 
 // NewPeeringHandler creates and initializes a new PeeringHandler with the
@@ -57,8 +65,9 @@ func NewPeeringHandler(db database.Store, logger *logger.Logger) *PeeringHandler
 	h := &PeeringHandler{
 		db:      db,
 		logger:  logger,
-		asPaths: make(map[string]time.Time),
+		asPaths: make(map[string]database.SeenTimes),
-		stopCh:  make(chan struct{}),
+		stopCh:  make(chan context.Context, 1),
+		done:    make(chan struct{}),
 	}
 
 	// Start the periodic processing goroutine
@@ -72,7 +81,7 @@ func NewPeeringHandler(db database.Store, logger *logger.Logger) *PeeringHandler
 // the AS path information needed to extract peering relationships.
 func (h *PeeringHandler) WantsMessage(messageType string) bool {
 	// We only care about UPDATE messages that have AS paths
-	return messageType == "UPDATE"
+	return messageType == ristypes.MessageTypeUpdate
 }
 
 // QueueCapacity returns the buffer size for the handler's message queue.
@@ -87,51 +96,41 @@ func (h *PeeringHandler) QueueCapacity() int {
 // than minPathLengthForPeering are ignored as they cannot contain valid
 // peering information.
 func (h *PeeringHandler) HandleMessage(msg *ristypes.RISMessage) {
+	// Peerings are between neighbours in the path's sequence only: the
+	// members of an AS set that ends the path are not neighbours of each
+	// other or of the AS before them.
+	sequence := msg.Path.Sequence
+
 	// Skip if no AS path or only one AS
-	if len(msg.Path) < minPathLengthForPeering {
+	if len(sequence) < minPathLengthForPeering {
 		return
 	}
 
 	timestamp := msg.ParsedTimestamp
 
 	// Encode AS path as JSON for use as map key
-	pathJSON, err := json.Marshal(msg.Path)
+	pathJSON, err := json.Marshal(sequence)
 	if err != nil {
 		h.logger.Error("Failed to encode AS path", "error", err)
 
 		return
 	}
+
 	key := string(pathJSON)
 
 	h.mu.Lock()
-	if _, exists := h.asPaths[key]; exists {
+	defer h.mu.Unlock()
-		// Already tracked: refresh its timestamp.
+
-		h.asPaths[key] = timestamp
+	seen, exists := h.asPaths[key]
-	} else if len(h.asPaths) >= maxTrackedPaths {
+	if !exists && len(h.asPaths) >= maxTrackedPaths {
 		// Map is full; drop this new path and count it.
 		h.droppedPaths++
-	} else {
-		h.asPaths[key] = timestamp
-	}
-	h.mu.Unlock()
-}
-
-// processLoop runs periodically to process AS paths into peerings
-func (h *PeeringHandler) processLoop() {
-	ticker := time.NewTicker(peeringProcessInterval)
-	defer ticker.Stop()
-
-	for {
-		select {
-		case <-ticker.C:
-			h.processPeerings()
-		case <-h.stopCh:
-			// Final processing
-			h.processPeerings()
 
 		return
 	}
-	}
+
+	seen.Add(timestamp)
+	h.asPaths[key] = seen
 }
 
 // ProcessPeeringsNow triggers immediate processing of all accumulated AS
@@ -141,15 +140,64 @@ func (h *PeeringHandler) ProcessPeeringsNow() {
 	h.processPeerings()
 }
 
+// Stop stops the processing goroutine, which writes the peerings of the AS
+// paths the handler holds before it returns, and waits for it until ctx ends.
+// Once ctx has ended the processing goroutine skips that last write, and a
+// write still running is lost when the process exits, so Stop logs how many
+// paths' peerings are not written.
+func (h *PeeringHandler) Stop(ctx context.Context) {
+	h.stopCh <- ctx
+
+	select {
+	case <-h.done:
+	case <-ctx.Done():
+	}
+
+	h.mu.Lock()
+	unwritten := h.writing + len(h.asPaths)
+	h.mu.Unlock()
+
+	if unwritten > 0 {
+		h.logger.Error("Stop deadline passed before the peering handler wrote "+
+			"its peerings", "paths", unwritten)
+	}
+}
+
+// processLoop runs periodically to process AS paths into peerings, and once
+// more on Stop.
+func (h *PeeringHandler) processLoop() {
+	defer close(h.done)
+
+	ticker := time.NewTicker(peeringProcessInterval)
+	defer ticker.Stop()
+
+	for {
+		select {
+		case <-ticker.C:
+			h.processPeerings()
+		case ctx := <-h.stopCh:
+			// Final processing, unless the stop deadline has passed
+			if ctx.Err() == nil {
+				h.processPeerings()
+			}
+
+			return
+		}
+	}
+}
+
 // processPeerings extracts peerings from AS paths and writes to database
 func (h *PeeringHandler) processPeerings() {
 	// Take the accumulated paths and replace the map with a fresh empty one
 	// under the lock. Each path is processed exactly once and the memory is
-	// released, so the map never grows past a single interval's traffic.
+	// released, so the map never grows past a single interval's traffic. The
+	// count of dropped paths starts again with the new map.
 	h.mu.Lock()
 	paths := h.asPaths
-	h.asPaths = make(map[string]time.Time)
+	h.asPaths = make(map[string]database.SeenTimes)
+	h.writing = len(paths)
 	dropped := h.droppedPaths
+	h.droppedPaths = 0
 	h.mu.Unlock()
 
 	if len(paths) == 0 {
@@ -160,11 +208,18 @@ func (h *PeeringHandler) processPeerings() {
 	type peeringKey struct {
 		low, high int
 	}
-	peerings := make(map[peeringKey]time.Time)
 
-	for pathJSON, timestamp := range paths {
+	// The earliest and the latest time an update carried each peering. Several
+	// paths, each carried at several times, can hold a peering, so a new
+	// peering is stored first seen at the earliest of them and last seen at the
+	// latest.
+	peerings := make(map[peeringKey]database.SeenTimes)
+
+	for pathJSON, pathSeen := range paths {
 		var path []int
-		if err := json.Unmarshal([]byte(pathJSON), &path); err != nil {
+
+		err := json.Unmarshal([]byte(pathJSON), &path)
+		if err != nil {
 			h.logger.Error("Failed to decode AS path", "error", err)
 
 			continue
@@ -187,18 +242,20 @@ func (h *PeeringHandler) processPeerings() {
 			}
 
 			key := peeringKey{low: low, high: high}
-			// Update timestamp if this is newer
+
-			if existing, ok := peerings[key]; !ok || timestamp.After(existing) {
+			seen := peerings[key]
-				peerings[key] = timestamp
+			seen.Add(pathSeen.First)
-			}
+			seen.Add(pathSeen.Last)
+			peerings[key] = seen
 		}
 	}
 
 	// Record peerings in database
 	start := time.Now()
 	successCount := 0
-	for key, ts := range peerings {
+
-		err := h.db.RecordPeering(key.low, key.high, ts)
+	for key, seen := range peerings {
+		err := h.db.RecordPeering(key.low, key.high, seen)
 		if err != nil {
 			h.logger.Error("Failed to record peering",
 				"as_a", key.low,
@@ -210,6 +267,10 @@ func (h *PeeringHandler) processPeerings() {
 		}
 	}
 
+	h.mu.Lock()
+	h.writing = 0
+	h.mu.Unlock()
+
 	h.logger.Info("Processed AS peerings",
 		"paths", len(paths),
 		"unique_peerings", len(peerings),
@@ -218,13 +279,3 @@ func (h *PeeringHandler) processPeerings() {
 		"duration", time.Since(start),
 	)
 }
-
-// Stop gracefully shuts down the handler by signaling the background
-// goroutine to stop and performing a final synchronous processing of
-// any remaining AS paths. This ensures no peering data is lost during
-// shutdown.
-func (h *PeeringHandler) Stop() {
-	close(h.stopCh)
-	// Process any remaining peerings synchronously
-	h.processPeerings()
-}

internal/routewatch/peeringhandler_internal_test.go

@@ -0,0 +1,374 @@
+package routewatch
+
+import (
+	"encoding/json"
+	"fmt"
+	"maps"
+	"slices"
+	"strconv"
+	"sync"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/ristypes"
+)
+
+const (
+	testASNA = 64500
+	testASNB = 64501
+	testASNC = 64502
+	testASND = 64503
+)
+
+// recordingStore wraps mockStore to count every RecordPeering call, so a
+// test can tell how many times a peering was written across separate runs.
+type recordingStore struct {
+	*mockStore
+
+	mu    sync.Mutex
+	calls int
+}
+
+func (r *recordingStore) RecordPeering(
+	asA, asB int, seen database.SeenTimes,
+) error {
+	r.mu.Lock()
+	r.calls++
+	r.mu.Unlock()
+
+	return r.mockStore.RecordPeering(asA, asB, seen)
+}
+
+func (r *recordingStore) callCount() int {
+	r.mu.Lock()
+	defer r.mu.Unlock()
+
+	return r.calls
+}
+
+// newTestHandler builds a PeeringHandler without starting the periodic
+// processing goroutine, so tests drive processing explicitly.
+func newTestHandler(db database.Store) *PeeringHandler {
+	return &PeeringHandler{
+		db:      db,
+		logger:  logger.New(),
+		asPaths: make(map[string]database.SeenTimes),
+	}
+}
+
+// snapshot returns how many paths the handler tracks and how many it dropped.
+func snapshot(h *PeeringHandler) (int, int) {
+	h.mu.Lock()
+	defer h.mu.Unlock()
+
+	return len(h.asPaths), h.droppedPaths
+}
+
+func pathKey(t *testing.T, asns ...int) string {
+	t.Helper()
+
+	b, err := json.Marshal(asns)
+	if err != nil {
+		t.Fatalf("failed to marshal path: %v", err)
+	}
+
+	return string(b)
+}
+
+func handle(h *PeeringHandler, ts time.Time, asns ...int) {
+	h.HandleMessage(&ristypes.RISMessage{
+		Path:            ristypes.ASPath{Sequence: asns},
+		ParsedTimestamp: ts,
+	})
+}
+
+// decodePath decodes an AS path from the JSON the feed writes for it, in which
+// an AS set is an array inside the path: decodePath(t, 1, 2, []int{3, 4})
+// decodes [1, 2, [3, 4]].
+func decodePath(t *testing.T, elements ...any) ristypes.ASPath {
+	t.Helper()
+
+	data, err := json.Marshal(elements)
+	if err != nil {
+		t.Fatalf("failed to marshal path: %v", err)
+	}
+
+	var path ristypes.ASPath
+
+	err = json.Unmarshal(data, &path)
+	if err != nil {
+		t.Fatalf("failed to decode path %s: %v", data, err)
+	}
+
+	return path
+}
+
+// TestPeeringHandlerLeavesATrailingASSetOutOfThePeerings handles AS paths that
+// end in an AS set. The members of a set are not neighbours of each other or of
+// the AS before the set, so a set of several members is in no peering; a set of
+// one member counts as that AS.
+func TestPeeringHandlerLeavesATrailingASSetOutOfThePeerings(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name string
+		path []any
+		// The peerings recorded, keyed as the mock store keys them.
+		wantPeerings []string
+	}{
+		{
+			name:         "a set of several members",
+			path:         []any{testASNA, testASNB, []int{testASNC, testASND}},
+			wantPeerings: []string{fmt.Sprintf("%d_%d", testASNA, testASNB)},
+		},
+		{
+			name: "a set of one member",
+			path: []any{testASNA, testASNB, []int{testASNC}},
+			wantPeerings: []string{
+				fmt.Sprintf("%d_%d", testASNA, testASNB),
+				fmt.Sprintf("%d_%d", testASNB, testASNC),
+			},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			store := newMockStore()
+			h := newTestHandler(store)
+			h.HandleMessage(&ristypes.RISMessage{
+				Path:            decodePath(t, tt.path...),
+				ParsedTimestamp: time.Now().UTC(),
+			})
+			h.ProcessPeeringsNow()
+
+			got := slices.Sorted(maps.Keys(store.Peerings))
+			if !slices.Equal(got, tt.wantPeerings) {
+				t.Errorf("peerings recorded = %v, want %v", got, tt.wantPeerings)
+			}
+		})
+	}
+}
+
+// TestPeeringHandlerRecordsNothingFromAPathWithASZero handles AS paths that
+// hold AS 0. RFC 7607 makes such a path malformed, so no peering is recorded
+// from it, not even one between two other ASes of the path.
+func TestPeeringHandlerRecordsNothingFromAPathWithASZero(t *testing.T) {
+	t.Parallel()
+
+	tests := []struct {
+		name string
+		path []any
+	}{
+		{name: "AS 0 between two ASes", path: []any{testASNA, 0, testASNB}},
+		{
+			name: "AS 0 after two ASes",
+			path: []any{testASNA, testASNB, 0, testASNC},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			store := newMockStore()
+			h := newTestHandler(store)
+			h.HandleMessage(&ristypes.RISMessage{
+				Path:            decodePath(t, tt.path...),
+				ParsedTimestamp: time.Now().UTC(),
+			})
+			h.ProcessPeeringsNow()
+
+			if len(store.Peerings) > 0 {
+				t.Errorf("peerings recorded = %v, want none",
+					slices.Sorted(maps.Keys(store.Peerings)))
+			}
+		})
+	}
+}
+
+// TestPeeringHandlerProcessesEachRunAndEmpties verifies that a processing run
+// empties the path map (the swap) and that a path seen again after a run is
+// recorded in the next run too.
+func TestPeeringHandlerProcessesEachRunAndEmpties(t *testing.T) {
+	t.Parallel()
+
+	store := &recordingStore{mockStore: newMockStore()}
+	h := newTestHandler(store)
+
+	now := time.Now().UTC()
+
+	// Run 1: one path, one peering recorded, map emptied afterwards.
+	handle(h, now, testASNA, testASNB)
+	h.ProcessPeeringsNow()
+
+	if tracked, _ := snapshot(h); tracked != 0 {
+		t.Fatalf("map not empty after first run: %d paths remain", tracked)
+	}
+
+	if got := store.callCount(); got != 1 {
+		t.Fatalf("want 1 RecordPeering call after first run, got %d", got)
+	}
+
+	// Run 2: the same path again is recorded again (RecordPeering upserts).
+	handle(h, now.Add(time.Second), testASNA, testASNB)
+	h.ProcessPeeringsNow()
+
+	if tracked, _ := snapshot(h); tracked != 0 {
+		t.Fatalf("map not empty after second run: %d paths remain", tracked)
+	}
+
+	if got := store.callCount(); got != 2 {
+		t.Fatalf("want 2 RecordPeering calls after second run, got %d", got)
+	}
+}
+
+// TestPeeringHandlerStopWritesTheRemainingPaths hands a running handler one
+// path and stops it. The peering must be in the database when Stop returns.
+func TestPeeringHandlerStopWritesTheRemainingPaths(t *testing.T) {
+	t.Parallel()
+
+	store := &recordingStore{mockStore: newMockStore()}
+	h := NewPeeringHandler(store, logger.New())
+
+	handle(h, time.Now().UTC(), testASNA, testASNB)
+	h.Stop(t.Context())
+
+	if got := store.callCount(); got != 1 {
+		t.Errorf("want 1 RecordPeering call when Stop returns, got %d", got)
+	}
+}
+
+// TestPeeringHandlerCapDropsAndCounts verifies that a full map drops new paths
+// and counts them, while a path already tracked is refreshed rather than
+// dropped.
+func TestPeeringHandlerCapDropsAndCounts(t *testing.T) {
+	t.Parallel()
+
+	store := &recordingStore{mockStore: newMockStore()}
+	h := newTestHandler(store)
+
+	now := time.Now().UTC()
+
+	// Fill the map to exactly maxTrackedPaths, including one real path key so
+	// the "already tracked" branch can be exercised. The filler keys are never
+	// processed in this test, so their contents do not matter.
+	existing := pathKey(t, testASNA, testASNB)
+
+	h.mu.Lock()
+	h.asPaths[existing] = database.SeenTimes{First: now, Last: now}
+
+	for i := 0; len(h.asPaths) < maxTrackedPaths; i++ {
+		h.asPaths[strconv.Itoa(i)] = database.SeenTimes{First: now, Last: now}
+	}
+	h.mu.Unlock()
+
+	// A new path is dropped and counted because the map is full.
+	handle(h, now.Add(time.Second), testASNA, testASNC)
+
+	tracked, dropped := snapshot(h)
+	if tracked != maxTrackedPaths {
+		t.Fatalf("want map size %d after drop, got %d", maxTrackedPaths, tracked)
+	}
+
+	if dropped != 1 {
+		t.Fatalf("want dropped count 1, got %d", dropped)
+	}
+
+	// A path already tracked is refreshed, not dropped.
+	refreshed := now.Add(2 * time.Second)
+	handle(h, refreshed, testASNA, testASNB)
+
+	tracked, dropped = snapshot(h)
+	if tracked != maxTrackedPaths {
+		t.Fatalf("want map size %d after refresh, got %d", maxTrackedPaths, tracked)
+	}
+
+	if dropped != 1 {
+		t.Fatalf("want dropped count still 1 after refresh, got %d", dropped)
+	}
+
+	h.mu.Lock()
+	gotSeen := h.asPaths[existing]
+	h.mu.Unlock()
+
+	if !gotSeen.Last.Equal(refreshed) {
+		t.Fatalf("existing path last seen not refreshed: want %v, got %v",
+			refreshed, gotSeen.Last)
+	}
+}
+
+// TestPeeringHandlerRunResetsTheDroppedCount fills the map, has the handler
+// drop one path, and runs the processing. The count of dropped paths is the
+// run's own, so it is 0 again after the run.
+func TestPeeringHandlerRunResetsTheDroppedCount(t *testing.T) {
+	t.Parallel()
+
+	h := newTestHandler(newMockStore())
+
+	now := time.Now().UTC()
+
+	// Paths of one AS carry no peering, so the run records nothing.
+	h.mu.Lock()
+	for asn := 1; len(h.asPaths) < maxTrackedPaths; asn++ {
+		h.asPaths[pathKey(t, asn)] = database.SeenTimes{First: now, Last: now}
+	}
+	h.mu.Unlock()
+
+	handle(h, now, testASNA, testASNB)
+
+	if _, dropped := snapshot(h); dropped != 1 {
+		t.Fatalf("want dropped count 1 with the map full, got %d", dropped)
+	}
+
+	h.ProcessPeeringsNow()
+
+	if _, dropped := snapshot(h); dropped != 0 {
+		t.Errorf("want dropped count 0 after a run, got %d", dropped)
+	}
+}
+
+// TestPeeringHandlerStoresTheEarliestAndLatestTimesOfANewPeering handles two
+// updates whose path carries a peering the database does not hold yet, the
+// second 5 seconds after the first, in one run into a real database. The
+// peering is stored first seen at the time of the first update and last seen
+// at the time of the second.
+func TestPeeringHandlerStoresTheEarliestAndLatestTimesOfANewPeering(
+	t *testing.T,
+) {
+	t.Parallel()
+
+	db := newTestDatabase(t)
+	h := newTestHandler(db)
+
+	first := peerAnnouncement(testPeerIP, testV4Prefix)
+	second := peerAnnouncement(testPeerIP, testV4Prefix)
+	second.ParsedTimestamp = first.ParsedTimestamp.Add(secondUpdateAfter)
+
+	h.HandleMessage(first)
+	h.HandleMessage(second)
+	h.ProcessPeeringsNow()
+
+	peers, err := db.GetASPeers(testASNA)
+	if err != nil {
+		t.Fatalf("GetASPeers(%d): %v", testASNA, err)
+	}
+
+	if len(peers) != 1 || peers[0].ASN != testASNB {
+		t.Fatalf("peers of AS%d = %+v, want AS%d alone", testASNA, peers,
+			testASNB)
+	}
+
+	if !peers[0].FirstSeen.Equal(first.ParsedTimestamp) {
+		t.Errorf("first_seen of peering AS%d-AS%d = %s, want %s", testASNA,
+			testASNB, peers[0].FirstSeen, first.ParsedTimestamp)
+	}
+
+	if !peers[0].LastSeen.Equal(second.ParsedTimestamp) {
+		t.Errorf("last_seen of peering AS%d-AS%d = %s, want %s", testASNA,
+			testASNB, peers[0].LastSeen, second.ParsedTimestamp)
+	}
+}

internal/routewatch/peeringhandler_test.go

@@ -1,163 +0,0 @@
-package routewatch
-
-import (
-	"encoding/json"
-	"strconv"
-	"sync"
-	"testing"
-	"time"
-
-	"git.eeqj.de/sneak/routewatch/internal/database"
-	"git.eeqj.de/sneak/routewatch/internal/logger"
-	"git.eeqj.de/sneak/routewatch/internal/ristypes"
-)
-
-const (
-	testASNA = 64500
-	testASNB = 64501
-	testASNC = 64502
-)
-
-// recordingStore wraps mockStore to count every RecordPeering call, so a
-// test can tell how many times a peering was written across separate runs.
-type recordingStore struct {
-	*mockStore
-
-	mu    sync.Mutex
-	calls int
-}
-
-func (r *recordingStore) RecordPeering(asA, asB int, ts time.Time) error {
-	r.mu.Lock()
-	r.calls++
-	r.mu.Unlock()
-
-	return r.mockStore.RecordPeering(asA, asB, ts)
-}
-
-func (r *recordingStore) callCount() int {
-	r.mu.Lock()
-	defer r.mu.Unlock()
-
-	return r.calls
-}
-
-// newTestHandler builds a PeeringHandler without starting the periodic
-// processing goroutine, so tests drive processing explicitly.
-func newTestHandler(db database.Store) *PeeringHandler {
-	return &PeeringHandler{
-		db:      db,
-		logger:  logger.New(),
-		asPaths: make(map[string]time.Time),
-		stopCh:  make(chan struct{}),
-	}
-}
-
-func snapshot(h *PeeringHandler) (tracked, dropped int) {
-	h.mu.Lock()
-	defer h.mu.Unlock()
-
-	return len(h.asPaths), h.droppedPaths
-}
-
-func pathKey(t *testing.T, asns ...int) string {
-	t.Helper()
-
-	b, err := json.Marshal(ristypes.ASPath(asns))
-	if err != nil {
-		t.Fatalf("failed to marshal path: %v", err)
-	}
-
-	return string(b)
-}
-
-func handle(h *PeeringHandler, ts time.Time, asns ...int) {
-	h.HandleMessage(&ristypes.RISMessage{
-		Path:            ristypes.ASPath(asns),
-		ParsedTimestamp: ts,
-	})
-}
-
-// TestPeeringHandlerProcessesEachRunAndEmpties verifies that a processing run
-// empties the path map (the swap) and that a path seen again after a run is
-// recorded in the next run too.
-func TestPeeringHandlerProcessesEachRunAndEmpties(t *testing.T) {
-	store := &recordingStore{mockStore: newMockStore()}
-	h := newTestHandler(store)
-
-	now := time.Now().UTC()
-
-	// Run 1: one path, one peering recorded, map emptied afterwards.
-	handle(h, now, testASNA, testASNB)
-	h.ProcessPeeringsNow()
-
-	if tracked, _ := snapshot(h); tracked != 0 {
-		t.Fatalf("map not empty after first run: %d paths remain", tracked)
-	}
-	if got := store.callCount(); got != 1 {
-		t.Fatalf("want 1 RecordPeering call after first run, got %d", got)
-	}
-
-	// Run 2: the same path again is recorded again (RecordPeering upserts).
-	handle(h, now.Add(time.Second), testASNA, testASNB)
-	h.ProcessPeeringsNow()
-
-	if tracked, _ := snapshot(h); tracked != 0 {
-		t.Fatalf("map not empty after second run: %d paths remain", tracked)
-	}
-	if got := store.callCount(); got != 2 {
-		t.Fatalf("want 2 RecordPeering calls after second run, got %d", got)
-	}
-}
-
-// TestPeeringHandlerCapDropsAndCounts verifies that a full map drops new paths
-// and counts them, while a path already tracked is refreshed rather than
-// dropped.
-func TestPeeringHandlerCapDropsAndCounts(t *testing.T) {
-	store := &recordingStore{mockStore: newMockStore()}
-	h := newTestHandler(store)
-
-	now := time.Now().UTC()
-
-	// Fill the map to exactly maxTrackedPaths, including one real path key so
-	// the "already tracked" branch can be exercised. The filler keys are never
-	// processed in this test, so their contents do not matter.
-	existing := pathKey(t, testASNA, testASNB)
-
-	h.mu.Lock()
-	h.asPaths[existing] = now
-	for i := 0; len(h.asPaths) < maxTrackedPaths; i++ {
-		h.asPaths[strconv.Itoa(i)] = now
-	}
-	h.mu.Unlock()
-
-	// A new path is dropped and counted because the map is full.
-	handle(h, now.Add(time.Second), testASNA, testASNC)
-
-	tracked, dropped := snapshot(h)
-	if tracked != maxTrackedPaths {
-		t.Fatalf("want map size %d after drop, got %d", maxTrackedPaths, tracked)
-	}
-	if dropped != 1 {
-		t.Fatalf("want dropped count 1, got %d", dropped)
-	}
-
-	// A path already tracked is refreshed, not dropped.
-	refreshed := now.Add(2 * time.Second)
-	handle(h, refreshed, testASNA, testASNB)
-
-	tracked, dropped = snapshot(h)
-	if tracked != maxTrackedPaths {
-		t.Fatalf("want map size %d after refresh, got %d", maxTrackedPaths, tracked)
-	}
-	if dropped != 1 {
-		t.Fatalf("want dropped count still 1 after refresh, got %d", dropped)
-	}
-
-	h.mu.Lock()
-	gotTS := h.asPaths[existing]
-	h.mu.Unlock()
-	if !gotTS.Equal(refreshed) {
-		t.Fatalf("existing path timestamp not refreshed: want %v, got %v", refreshed, gotTS)
-	}
-}

internal/routewatch/prefixhandler.go

@@ -1,8 +1,10 @@
 package routewatch
 
 import (
+	"context"
 	"net"
 	"net/netip"
+	"slices"
 	"strings"
 	"sync"
 	"time"
@@ -11,41 +13,94 @@ import (
 	"git.eeqj.de/sneak/routewatch/internal/logger"
 	"git.eeqj.de/sneak/routewatch/internal/metrics"
 	"git.eeqj.de/sneak/routewatch/internal/ristypes"
-	"github.com/google/uuid"
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
 )
 
 const (
 	// prefixHandlerQueueSize is the queue capacity for prefix tracking
-	// operations, about 4 seconds of feed at peak. The streamer drops rather
+	// operations, about 4 seconds of feed at peak. The streamer drops updates
-	// than blocks when a queue is full, so this bounds memory. Batches still
+	// rather than blocks when a queue is full (a STATE message is held until
-	// flush on a timer (prefixBatchTimeout).
+	// there is room), so this bounds memory. Batches still flush on a timer
+	// (prefixBatchTimeout).
 	prefixHandlerQueueSize = 20000
 
 	// prefixBatchSize is the number of prefix updates to batch together
 	prefixBatchSize = 25000
 
+	// prefixBatchMaxSize is the most prefix updates the batch holds while the
+	// flush goroutine writes the previous one: ten batches, about 60 MiB
+	// (README "Memory"). At this size HandleMessage waits until the flush
+	// goroutine takes the batch. On top of these, the handler holds at most
+	// prefixBatchSize updates, about 6 MiB, that a flush kept for the peers
+	// whose route delete failed (held).
+	prefixBatchMaxSize = 250000
+
 	// prefixBatchTimeout is the maximum time to wait before flushing a batch
 	// DO NOT reduce this timeout - larger batches are more efficient
 	prefixBatchTimeout = 1 * time.Second
 
+	// maxTableChangesPerWrite is the most changed prefix and origin AS pairs
+	// of the routing table that one flush writes; the rest wait for the next
+	// flushes. The change of one pair takes about 430 bytes of memory and 15
+	// microseconds to write. So a write holds about 9 MB and takes about a
+	// third of the one-second flush interval, which leaves the rest to the
+	// batch. The feed peaks at about 5,000 updates a second, each changing at
+	// most one pair. So the million pairs that a full-table peer's session
+	// going down changes are written in about a minute.
+	maxTableChangesPerWrite = 20000
+
 	// IP version constants
 	ipv4Version = 4
 	ipv6Version = 6
 )
 
-// PrefixHandler tracks BGP prefixes and maintains a live routing table in the database.
+// PrefixHandler tracks BGP prefixes and maintains a live routing table in the
-// Routes are added on announcement and deleted on withdrawal.
+// database. Routes are added on announcement and deleted on withdrawal or on an
+// announcement whose AS path is empty or holds AS 0 anywhere, and all routes of
+// a peer are deleted when its BGP session goes down.
+//
+// It applies the same updates to the routing table in memory as it handles
+// them, and writes the table's changes to the database with each flush.
 type PrefixHandler struct {
 	db      database.Store
 	logger  *logger.Logger
 	metrics *metrics.Tracker
+	table   *routetable.Table
 
-	// Batching
+	// Batching. mu guards batch, held, peersDown, writing, deleting and
+	// writingTableChanges only: the flush goroutine takes them under mu and
+	// writes them after releasing mu, so HandleMessage waits for a database
+	// write only while batch is at prefixBatchMaxSize.
 	mu    sync.Mutex
 	batch []prefixUpdate
-	lastFlush time.Time
+	// held holds the updates that the peers whose route delete failed in the
-	stopCh    chan struct{}
+	// last flush sent after their session went down, for the next flush to
-	wg        sync.WaitGroup
+	// write after it retries the delete. They are not in batch, so they do
+	// not count toward a full batch: a delete that keeps failing is retried
+	// at the pace of the flush timer.
+	held []prefixUpdate
+	// peersDown holds, each once, the peers whose BGP session went down
+	// since the batch was taken, and those whose route delete failed in the
+	// last flush.
+	peersDown []string
+	// writing is how many updates the flush goroutine took and has not
+	// finished writing.
+	writing int
+	// deleting is how many peers from peersDown the flush goroutine took and
+	// has not yet deleted the routes of.
+	deleting int
+	// writingTableChanges is how many changes the flush goroutine took from
+	// the routing table and has not finished writing.
+	writingTableChanges int
+	// batchTaken is signalled when the flush goroutine takes batch; its L is
+	// &mu.
+	batchTaken  sync.Cond
+	batchFullCh chan struct{} // wakes the flush goroutine when batch is full
+	lastFlush   time.Time     // used only by the flush goroutine
+	// stopCh hands Stop's context to the flush goroutine. It holds one, so
+	// Stop does not wait for a write in progress to hand it over.
+	stopCh chan context.Context
+	done   chan struct{} // closed when the flush goroutine returns
 }
 
 type prefixUpdate struct {
@@ -55,20 +110,40 @@ type prefixUpdate struct {
 	messageType string
 	timestamp   time.Time
 	path        []int
+	// endsInSet says the path ends in an AS set of several members, so it
+	// gives no origin.
+	endsInSet bool
 }
 
-// NewPrefixHandler creates a new batched prefix tracking handler
+// announcesRoute says whether u gives the peer a route to the prefix: an
-func NewPrefixHandler(db database.Store, logger *logger.Logger) *PrefixHandler {
+// announcement whose path gives an origin, or ends in an AS set of several
+// members, which gives origin 0 (none). Any other update removes the peer's
+// route for the prefix: a withdrawal, or an announcement whose path is empty
+// or holds AS 0 anywhere (the decoder empties such a path). Such an
+// announcement replaces the peer's route, as any announcement does, with none
+// (RFC 7607 treats AS 0 in a path as a withdrawal).
+func (u prefixUpdate) announcesRoute() bool {
+	return u.messageType == "announcement" && (u.originASN > 0 || u.endsInSet)
+}
+
+// NewPrefixHandler creates a new batched prefix tracking handler, which applies
+// the updates to table, the routing table read from db at startup.
+func NewPrefixHandler(
+	db database.Store, logger *logger.Logger, table *routetable.Table,
+) *PrefixHandler {
 	h := &PrefixHandler{
 		db:          db,
 		logger:      logger,
+		table:       table,
 		batch:       make([]prefixUpdate, 0, prefixBatchSize),
+		batchFullCh: make(chan struct{}, 1),
 		lastFlush:   time.Now(),
-		stopCh:    make(chan struct{}),
+		stopCh:      make(chan context.Context, 1),
+		done:        make(chan struct{}),
 	}
+	h.batchTaken.L = &h.mu
 
 	// Start the flush timer goroutine
-	h.wg.Add(1)
 	go h.flushLoop()
 
 	return h
@@ -79,10 +154,13 @@ func (h *PrefixHandler) SetMetricsTracker(metrics *metrics.Tracker) {
 	h.metrics = metrics
 }
 
-// WantsMessage returns true if this handler wants to process messages of the given type
+// WantsMessage returns true if this handler wants to process messages of the
+// given type
 func (h *PrefixHandler) WantsMessage(messageType string) bool {
-	// We only care about UPDATE messages for the routing table
+	// UPDATE messages change routes; a STATE message can say a peer's
-	return messageType == "UPDATE"
+	// session went down, which removes all of its routes.
+	return messageType == ristypes.MessageTypeUpdate ||
+		messageType == ristypes.MessageTypeState
 }
 
 // QueueCapacity returns the desired queue capacity for this handler
@@ -93,29 +171,69 @@ func (h *PrefixHandler) QueueCapacity() int {
 
 // HandleMessage processes a message to track prefix information
 func (h *PrefixHandler) HandleMessage(msg *ristypes.RISMessage) {
+	if msg.Type == ristypes.MessageTypeState {
+		if msg.State == ristypes.StateDown {
+			h.peerSessionDown(msg.Peer)
+		}
+
+		return
+	}
+
 	// Use the pre-parsed timestamp
 	timestamp := msg.ParsedTimestamp
 
-	// Get origin ASN from path (last element)
+	originASN := msg.Path.Origin()
-	var originASN int
+
-	if len(msg.Path) > 0 {
+	// The stored and shown AS path lists the members of an AS set that ends
-		originASN = msg.Path[len(msg.Path)-1]
+	// the path after the rest of it.
-	}
+	path := slices.Concat(msg.Path.Sequence, msg.Path.Set)
 
 	h.mu.Lock()
 	defer h.mu.Unlock()
 
-	// Process announcements
+	// While the batch is at its maximum size, wait until the flush goroutine
+	// takes it. Meanwhile the handler takes nothing off its queue, which
+	// fills, and the streamer drops messages for it.
+	for len(h.batch) >= prefixBatchMaxSize {
+		h.batchTaken.Wait()
+	}
+
+	// Withdrawals are queued, and applied to the routing table, before
+	// announcements, because the batch keeps each peer's last update for a
+	// prefix: a message that lists a prefix as both withdrawn and announced
+	// keeps the route (RFC 4271 section 4.3). A withdrawal names only the
+	// prefix and peer: the message's AS path, and so its origin AS, belongs
+	// to the announced prefixes.
+	for _, prefix := range msg.Withdrawals {
+		update := prefixUpdate{
+			prefix:      canonicalPrefix(prefix),
+			peer:        msg.Peer,
+			messageType: "withdrawal",
+			timestamp:   timestamp,
+		}
+		h.batch = append(h.batch, update)
+		h.applyToTable(update)
+
+		// Record withdrawal in metrics
+		if h.metrics != nil {
+			h.metrics.RecordWithdrawal()
+		}
+	}
+
 	for _, announcement := range msg.Announcements {
 		for _, prefix := range announcement.Prefixes {
-			h.batch = append(h.batch, prefixUpdate{
+			update := prefixUpdate{
 				prefix:      canonicalPrefix(prefix),
 				originASN:   originASN,
 				peer:        msg.Peer,
 				messageType: "announcement",
 				timestamp:   timestamp,
-				path:        msg.Path,
+				path:        path,
-			})
+				endsInSet:   len(msg.Path.Set) > 0,
+			}
+			h.batch = append(h.batch, update)
+			h.applyToTable(update)
+
 			// Record announcement in metrics
 			if h.metrics != nil {
 				h.metrics.RecordAnnouncement()
@@ -123,96 +241,197 @@ func (h *PrefixHandler) HandleMessage(msg *ristypes.RISMessage) {
 		}
 	}
 
-	// Process withdrawals
+	// A full batch wakes the flush goroutine. If it is writing, the wake-up
-	for _, prefix := range msg.Withdrawals {
+	// waits in batchFullCh and the batch keeps growing until the write ends
-		h.batch = append(h.batch, prefixUpdate{
+	// or the batch reaches prefixBatchMaxSize.
-			prefix:      canonicalPrefix(prefix),
-			originASN:   originASN, // Use the originASN from path if available
-			peer:        msg.Peer,
-			messageType: "withdrawal",
-			timestamp:   timestamp,
-			path:        msg.Path,
-		})
-		// Record withdrawal in metrics
-		if h.metrics != nil {
-			h.metrics.RecordWithdrawal()
-		}
-	}
-
-	// Check if we need to flush
 	if len(h.batch) >= prefixBatchSize {
-		h.flushBatchLocked()
+		select {
+		case h.batchFullCh <- struct{}{}:
+		default: // already woken
+		}
 	}
 }
 
-// flushLoop runs in a goroutine and periodically flushes batches
+// Stop stops the flush goroutine, which makes one last flush before it returns,
+// and waits for it until ctx ends. That flush writes the updates and peer downs
+// the handler holds, and at most maxTableChangesPerWrite of the routing table's
+// changes. Once ctx has ended the flush goroutine skips that last flush, and a
+// write still running is lost when the process exits. A peer whose route delete
+// fails in the last flush is not retried either, nor is a failed write of the
+// table's changes. Stop logs how many updates are not written, how many peer
+// downs have not had their routes deleted, and how many of the table's changes
+// are not written.
+func (h *PrefixHandler) Stop(ctx context.Context) {
+	h.stopCh <- ctx
+
+	select {
+	case <-h.done:
+	case <-ctx.Done():
+	}
+
+	h.mu.Lock()
+	unwritten := h.writing + len(h.batch) + len(h.held)
+	peersDown := h.deleting + len(h.peersDown)
+	tableChanges := h.writingTableChanges + h.table.ChangeCount()
+	h.mu.Unlock()
+
+	if unwritten == 0 && peersDown == 0 && tableChanges == 0 {
+		return
+	}
+
+	if ctx.Err() != nil {
+		h.logger.Error("Stop deadline passed before the prefix handler wrote "+
+			"its updates", "updates", unwritten, "peers_down", peersDown,
+			"table_changes", tableChanges)
+
+		return
+	}
+
+	h.logger.Error("The prefix handler stopped before it wrote everything",
+		"updates", unwritten, "peers_down", peersDown,
+		"table_changes", tableChanges)
+}
+
+// peerSessionDown notes that the peer's BGP session went down, so that the next
+// flush deletes all of the peer's live routes before it writes the batch. The
+// peer's updates still in the batch or held came before the down, which
+// replaces them, so they are dropped; updates the peer sends later stay in the
+// batch and are written after the delete. A peer already in peersDown is not
+// added again: one delete removes all of its routes. In the routing table in
+// memory the down takes effect at once.
+func (h *PrefixHandler) peerSessionDown(peer string) {
+	h.mu.Lock()
+	defer h.mu.Unlock()
+
+	h.table.PeerDown(peer)
+
+	isPeers := func(update prefixUpdate) bool { return update.peer == peer }
+	h.batch = slices.DeleteFunc(h.batch, isPeers)
+	h.held = slices.DeleteFunc(h.held, isPeers)
+
+	if !slices.Contains(h.peersDown, peer) {
+		h.peersDown = append(h.peersDown, peer)
+	}
+}
+
+// flushLoop runs in a goroutine, the only one that writes the batch: it
+// flushes when the batch is full, periodically, and once more on Stop.
 func (h *PrefixHandler) flushLoop() {
-	defer h.wg.Done()
+	defer close(h.done)
+
 	ticker := time.NewTicker(prefixBatchTimeout)
 	defer ticker.Stop()
 
 	for {
 		select {
 		case <-ticker.C:
-			h.mu.Lock()
 			if time.Since(h.lastFlush) >= prefixBatchTimeout {
-				h.flushBatchLocked()
+				h.flushBatch()
+			}
+		case <-h.batchFullCh:
+			h.flushBatch()
+		case ctx := <-h.stopCh:
+			// Final flush, unless the stop deadline has passed
+			if ctx.Err() == nil {
+				h.flushBatch()
 			}
-			h.mu.Unlock()
-		case <-h.stopCh:
-			// Final flush
-			h.mu.Lock()
-			h.flushBatchLocked()
-			h.mu.Unlock()
 
 			return
 		}
 	}
 }
 
-// flushBatchLocked flushes the prefix batch to the database (must be called with mutex held)
+// takeBatch returns the pending updates, the held ones first, and the peers in
-func (h *PrefixHandler) flushBatchLocked() {
+// peersDown, and leaves an empty batch, no held updates and no peers in their
-	if len(h.batch) == 0 {
+// place for HandleMessage to fill. The updates count in writing until
+// flushBatch has written them, and the peers in deleting until their routes
+// are deleted.
+func (h *PrefixHandler) takeBatch() ([]prefixUpdate, []string) {
+	h.mu.Lock()
+	defer h.mu.Unlock()
+
+	if len(h.held) == 0 && len(h.batch) == 0 && len(h.peersDown) == 0 {
+		return nil, nil
+	}
+
+	batch, peersDown := h.batch, h.peersDown
+	if len(h.held) > 0 {
+		batch = slices.Concat(h.held, h.batch)
+	}
+
+	h.writing = len(batch)
+	h.deleting = len(peersDown)
+	h.held = nil
+	h.batch = make([]prefixUpdate, 0, prefixBatchSize)
+	h.peersDown = nil
+	h.batchTaken.Broadcast()
+
+	return batch, peersDown
+}
+
+// flushBatch writes the routing table's changes, then takes the pending batch
+// and writes it to the database, without holding the mutex during the writes.
+func (h *PrefixHandler) flushBatch() {
+	h.writeTableChanges()
+
+	batch, peersDown := h.takeBatch()
+	if len(batch) == 0 && len(peersDown) == 0 {
 		return
 	}
 
 	startTime := time.Now()
-	batchSize := len(h.batch)
+	batchSize := len(batch)
 
-	// Group updates by prefix to deduplicate
+	// The routes of a peer whose session went down are deleted first: the
-	// For each prefix, keep the latest update
+	// batch holds only the updates the peer sent after the down. A peer whose
-	prefixMap := make(map[string]prefixUpdate)
+	// delete failed is kept, with those updates, for the next flush to retry.
-	for _, update := range h.batch {
+	failedPeers := h.deleteLiveRoutesOfPeers(peersDown)
-		key := update.prefix
+	if len(failedPeers) > 0 {
-		if existing, ok := prefixMap[key]; !ok || update.timestamp.After(existing.timestamp) {
+		batch = h.keepForNextFlush(batch, failedPeers)
-			prefixMap[key] = update
+	}
+
+	// For each prefix and peer, keep only the last update the feed delivered.
+	// Updates from different peers for the same prefix are all kept. Arrival
+	// order decides, not the timestamp: timestamps are whole seconds, so
+	// updates within one second carry the same one.
+	type prefixAndPeer struct{ prefix, peer string }
+
+	lastUpdates := make(map[prefixAndPeer]prefixUpdate)
+
+	// Each prefix the batch names, announced or withdrawn, with the earliest
+	// timestamp among all of its updates, not only each peer's last one: the
+	// prefix tables store only first_seen, written when the prefix is new and
+	// never changed after.
+	prefixesToUpdate := make(map[string]time.Time)
+
+	for _, update := range batch {
+		lastUpdates[prefixAndPeer{update.prefix, update.peer}] = update
+
+		seen, exists := prefixesToUpdate[update.prefix]
+		if !exists || update.timestamp.Before(seen) {
+			prefixesToUpdate[update.prefix] = update.timestamp
 		}
 	}
 
 	// Collect routes to upsert and delete
-	var routesToUpsert []*database.LiveRoute
+	var (
-	var routesToDelete []database.LiveRouteDeletion
+		routesToUpsert []*database.LiveRoute
+		routesToDelete []database.LiveRouteDeletion
+	)
 
-	// Collect unique prefixes to update
+	for _, update := range lastUpdates {
-	prefixesToUpdate := make(map[string]time.Time)
+		if update.announcesRoute() {
-
-	for _, update := range prefixMap {
-		// Track prefix for both announcements and withdrawals
-		if _, exists := prefixesToUpdate[update.prefix]; !exists || update.timestamp.After(prefixesToUpdate[update.prefix]) {
-			prefixesToUpdate[update.prefix] = update.timestamp
-		}
-
-		if update.messageType == "announcement" && update.originASN > 0 {
 			// Create live route for batch upsert
 			route := h.createLiveRoute(update)
 			if route != nil {
 				routesToUpsert = append(routesToUpsert, route)
 			}
-		} else if update.messageType == "withdrawal" {
+		} else {
 			// Parse CIDR to get IP version
 			_, ipVersion, err := parseCIDR(update.prefix)
 			if err != nil {
-				h.logger.Error("Failed to parse CIDR for withdrawal", "prefix", update.prefix, "error", err)
+				h.logger.Error("Failed to parse CIDR for route deletion",
+					"prefix", update.prefix, "error", err)
 
 				continue
 			}
@@ -220,26 +439,201 @@ func (h *PrefixHandler) flushBatchLocked() {
 			// Create deletion record for batch delete
 			routesToDelete = append(routesToDelete, database.LiveRouteDeletion{
 				Prefix:    update.prefix,
-				OriginASN: update.originASN,
 				PeerIP:    update.peer,
 				IPVersion: ipVersion,
 			})
 		}
 	}
 
-	// Process batch operations
+	successCount := h.writeBatch(routesToUpsert, routesToDelete, prefixesToUpdate)
+
+	h.mu.Lock()
+	h.writing = 0
+	h.mu.Unlock()
+
+	elapsed := time.Since(startTime)
+	h.logger.Debug("Flushed prefix batch",
+		"batch_size", batchSize,
+		"peers_down", len(peersDown),
+		"unique_prefixes", len(prefixesToUpdate),
+		"success", successCount,
+		"duration_ms", elapsed.Milliseconds(),
+	)
+
+	h.lastFlush = time.Now()
+}
+
+// applyToTable applies update to the routing table in memory: as an
+// announcement when it gives the peer a route, otherwise as a withdrawal. A
+// prefix that does not parse is left out; flushBatch logs it.
+func (h *PrefixHandler) applyToTable(update prefixUpdate) {
+	prefix, err := netip.ParsePrefix(update.prefix)
+	if err != nil {
+		return
+	}
+
+	if !update.announcesRoute() {
+		h.table.Withdraw(prefix, update.peer)
+
+		return
+	}
+
+	h.table.Announce(prefix, update.originASN, routetable.Route{
+		Peer:   update.peer,
+		ASPath: update.path,
+		Time:   update.timestamp,
+	})
+}
+
+// writeTableChanges writes at most maxTableChangesPerWrite changes of the
+// routing table in memory, and every peer the table numbered since the last
+// write.
+// When the write fails, they go back to the table, so that a later write writes
+// them again, each pair with its content then.
+func (h *PrefixHandler) writeTableChanges() {
+	h.mu.Lock()
+	changes, newPeers := h.table.TakeChanges(maxTableChangesPerWrite)
+	h.writingTableChanges = len(changes)
+	h.mu.Unlock()
+
+	if len(changes) == 0 && len(newPeers) == 0 {
+		return
+	}
+
+	err := h.db.WriteRouteChanges(changes, newPeers)
+
+	h.mu.Lock()
+	defer h.mu.Unlock()
+
+	h.writingTableChanges = 0
+
+	if err != nil {
+		h.logger.Error("Failed to write the routing table's changes; the next "+
+			"flush retries", "changes", len(changes), "new_peers", len(newPeers),
+			"error", err)
+
+		h.table.PutBackChanges(changes, newPeers)
+	}
+}
+
+// deleteLiveRoutesOfPeers deletes all live routes of each of peers, the peers
+// whose BGP session went down, and returns those whose delete failed. Most
+// downs come from peers that have no routes, so only a delete that removed
+// routes is logged. A peer whose delete succeeded no longer counts in deleting;
+// a failed one counts until keepForNextFlush puts it back in peersDown.
+func (h *PrefixHandler) deleteLiveRoutesOfPeers(peers []string) []string {
+	var failed []string
+
+	for _, peer := range peers {
+		deleted, err := h.db.DeleteLiveRoutesOfPeer(peer)
+		if err != nil {
+			h.logger.Error(
+				"Failed to delete the live routes of a peer whose session "+
+					"went down; the next flush retries",
+				"peer", peer, "routes_deleted", deleted, "error", err)
+
+			failed = append(failed, peer)
+
+			continue
+		}
+
+		h.mu.Lock()
+		h.deleting--
+		h.mu.Unlock()
+
+		if deleted > 0 {
+			h.logger.Info(
+				"Deleted the live routes of a peer whose session went down",
+				"peer", peer, "routes", deleted)
+		}
+	}
+
+	return failed
+}
+
+// keepForNextFlush puts the peers whose route delete failed back in peersDown,
+// and their updates from batch in held, so that the next flush retries each
+// delete before it writes the peer's updates; written now, they would be
+// deleted by the retry. The updates of a peer whose session went down again
+// since the batch was taken came before that down, so they are dropped, as
+// peerSessionDown drops them, and the peer is already in peersDown. It returns
+// the updates of the other peers, for this flush to write.
+//
+// While a delete keeps failing, the held updates would grow with every flush,
+// so only the last prefixBatchSize of them are kept and the older ones are
+// dropped. This never leaves in place a route the peer withdrew: a prefix with
+// a kept update keeps its last one, and the retry deletes all of the peer's
+// routes, so a prefix whose updates were all dropped has no route from the peer
+// until the peer announces it again.
+func (h *PrefixHandler) keepForNextFlush(
+	batch []prefixUpdate, failedPeers []string,
+) []prefixUpdate {
+	var kept, toWrite []prefixUpdate
+
+	for _, update := range batch {
+		if slices.Contains(failedPeers, update.peer) {
+			kept = append(kept, update)
+		} else {
+			toWrite = append(toWrite, update)
+		}
+	}
+
+	h.mu.Lock()
+	defer h.mu.Unlock()
+
+	kept = slices.DeleteFunc(kept, func(update prefixUpdate) bool {
+		return slices.Contains(h.peersDown, update.peer)
+	})
+
+	if len(kept) > prefixBatchSize {
+		h.logger.Error(
+			"Dropped the oldest updates of peers whose route delete failed",
+			"peers", failedPeers, "dropped", len(kept)-prefixBatchSize)
+
+		// A copy, so that the dropped updates are freed.
+		kept = slices.Clone(kept[len(kept)-prefixBatchSize:])
+	}
+
+	// takeBatch left held empty, and only this flush fills it. The kept
+	// updates now count in held, not in writing, and the failed peers in
+	// peersDown, not in deleting.
+	h.held = kept
+	h.writing = len(toWrite)
+	h.deleting = 0
+
+	for _, peer := range failedPeers {
+		if !slices.Contains(h.peersDown, peer) {
+			h.peersDown = append(h.peersDown, peer)
+		}
+	}
+
+	return toWrite
+}
+
+// writeBatch writes the collected route upserts, route deletions and prefix
+// updates to the database and returns how many routes it wrote.
+func (h *PrefixHandler) writeBatch(
+	routesToUpsert []*database.LiveRoute,
+	routesToDelete []database.LiveRouteDeletion,
+	prefixesToUpdate map[string]time.Time,
+) int {
 	successCount := 0
+
 	if len(routesToUpsert) > 0 {
-		if err := h.db.UpsertLiveRouteBatch(routesToUpsert); err != nil {
+		err := h.db.UpsertLiveRouteBatch(routesToUpsert)
-			h.logger.Error("Failed to upsert route batch", "error", err, "count", len(routesToUpsert))
+		if err != nil {
+			h.logger.Error("Failed to upsert route batch",
+				"error", err, "count", len(routesToUpsert))
 		} else {
 			successCount += len(routesToUpsert)
 		}
 	}
 
 	if len(routesToDelete) > 0 {
-		if err := h.db.DeleteLiveRouteBatch(routesToDelete); err != nil {
+		err := h.db.DeleteLiveRouteBatch(routesToDelete)
-			h.logger.Error("Failed to delete route batch", "error", err, "count", len(routesToDelete))
+		if err != nil {
+			h.logger.Error("Failed to delete route batch",
+				"error", err, "count", len(routesToDelete))
 		} else {
 			successCount += len(routesToDelete)
 		}
@@ -247,22 +641,14 @@ func (h *PrefixHandler) flushBatchLocked() {
 
 	// Update prefix tables
 	if len(prefixesToUpdate) > 0 {
-		if err := h.db.UpdatePrefixesBatch(prefixesToUpdate); err != nil {
+		err := h.db.UpdatePrefixesBatch(prefixesToUpdate)
-			h.logger.Error("Failed to update prefix batch", "error", err, "count", len(prefixesToUpdate))
+		if err != nil {
+			h.logger.Error("Failed to update prefix batch",
+				"error", err, "count", len(prefixesToUpdate))
 		}
 	}
 
-	elapsed := time.Since(startTime)
+	return successCount
-	h.logger.Debug("Flushed prefix batch",
-		"batch_size", batchSize,
-		"unique_prefixes", len(prefixMap),
-		"success", successCount,
-		"duration_ms", elapsed.Milliseconds(),
-	)
-
-	// Clear batch
-	h.batch = h.batch[:0]
-	h.lastFlush = time.Now()
 }
 
 // canonicalPrefix returns prefix in the text form net/netip prints, the form
@@ -280,8 +666,8 @@ func canonicalPrefix(prefix string) string {
 	return p.Masked().String()
 }
 
-// parseCIDR extracts the mask length and IP version from a prefix string
+// parseCIDR returns the mask length and the IP version of a prefix string
-func parseCIDR(prefix string) (maskLength int, ipVersion int, err error) {
+func parseCIDR(prefix string) (int, int, error) {
 	_, ipNet, err := net.ParseCIDR(prefix)
 	if err != nil {
 		return 0, 0, err
@@ -295,50 +681,6 @@ func parseCIDR(prefix string) (maskLength int, ipVersion int, err error) {
 	return ones, ipv4Version, nil
 }
 
-// processAnnouncement handles storing an announcement in the database
-// nolint:unused // kept for potential future use
-func (h *PrefixHandler) processAnnouncement(_ *database.Prefix, update prefixUpdate) {
-	// Parse CIDR to get mask length
-	maskLength, ipVersion, err := parseCIDR(update.prefix)
-	if err != nil {
-		h.logger.Error("Failed to parse CIDR",
-			"prefix", update.prefix,
-			"error", err,
-		)
-
-		return
-	}
-
-	// Track route update metrics
-	if h.metrics != nil {
-		if ipVersion == ipv4Version {
-			h.metrics.RecordIPv4Update()
-		} else {
-			h.metrics.RecordIPv6Update()
-		}
-	}
-
-	// Create live route record
-	liveRoute := &database.LiveRoute{
-		ID:          uuid.New(),
-		Prefix:      update.prefix,
-		MaskLength:  maskLength,
-		IPVersion:   ipVersion,
-		OriginASN:   update.originASN,
-		PeerIP:      update.peer,
-		ASPath:      update.path,
-		NextHop:     update.peer, // Using peer as next hop
-		LastUpdated: update.timestamp,
-	}
-
-	if err := h.db.UpsertLiveRoute(liveRoute); err != nil {
-		h.logger.Error("Failed to upsert live route",
-			"prefix", update.prefix,
-			"error", err,
-		)
-	}
-}
-
 // createLiveRoute creates a LiveRoute from a prefix update
 func (h *PrefixHandler) createLiveRoute(update prefixUpdate) *database.LiveRoute {
 	// Parse CIDR to get mask length
@@ -363,7 +705,6 @@ func (h *PrefixHandler) createLiveRoute(update prefixUpdate) *database.LiveRoute
 
 	// Create live route record
 	liveRoute := &database.LiveRoute{
-		ID:          uuid.New(),
 		Prefix:      update.prefix,
 		MaskLength:  maskLength,
 		IPVersion:   ipVersion,
@@ -376,104 +717,3 @@ func (h *PrefixHandler) createLiveRoute(update prefixUpdate) *database.LiveRoute
 
 	return liveRoute
 }
-
-// processAnnouncementDirect handles storing an announcement directly without prefix table
-// nolint:unused // kept for potential future use
-func (h *PrefixHandler) processAnnouncementDirect(update prefixUpdate) {
-	// Parse CIDR to get mask length
-	maskLength, ipVersion, err := parseCIDR(update.prefix)
-	if err != nil {
-		h.logger.Error("Failed to parse CIDR",
-			"prefix", update.prefix,
-			"error", err,
-		)
-
-		return
-	}
-
-	// Track route update metrics
-	if h.metrics != nil {
-		if ipVersion == ipv4Version {
-			h.metrics.RecordIPv4Update()
-		} else {
-			h.metrics.RecordIPv6Update()
-		}
-	}
-
-	// Create live route record
-	liveRoute := &database.LiveRoute{
-		ID:          uuid.New(),
-		Prefix:      update.prefix,
-		MaskLength:  maskLength,
-		IPVersion:   ipVersion,
-		OriginASN:   update.originASN,
-		PeerIP:      update.peer,
-		ASPath:      update.path,
-		NextHop:     update.peer, // Using peer as next hop
-		LastUpdated: update.timestamp,
-	}
-
-	if err := h.db.UpsertLiveRoute(liveRoute); err != nil {
-		h.logger.Error("Failed to upsert live route",
-			"prefix", update.prefix,
-			"error", err,
-		)
-	}
-}
-
-// processWithdrawalDirect handles removing a route directly without prefix table
-// nolint:unused // kept for potential future use
-func (h *PrefixHandler) processWithdrawalDirect(update prefixUpdate) {
-	// For withdrawals, we need to delete the route from live_routes
-	if update.originASN > 0 {
-		if err := h.db.DeleteLiveRoute(update.prefix, update.originASN, update.peer); err != nil {
-			h.logger.Error("Failed to delete live route",
-				"prefix", update.prefix,
-				"origin_asn", update.originASN,
-				"peer", update.peer,
-				"error", err,
-			)
-		}
-	} else {
-		// If no origin ASN, just delete all routes for this prefix from this peer
-		if err := h.db.DeleteLiveRoute(update.prefix, 0, update.peer); err != nil {
-			h.logger.Error("Failed to delete live route (no origin ASN)",
-				"prefix", update.prefix,
-				"peer", update.peer,
-				"error", err,
-			)
-		}
-	}
-}
-
-// processWithdrawal handles removing a route from the live routing table
-// nolint:unused // kept for potential future use
-func (h *PrefixHandler) processWithdrawal(_ *database.Prefix, update prefixUpdate) {
-	// For withdrawals, we need to delete the route from live_routes
-	// Since we have the origin ASN from the update, we can delete the specific route
-	if update.originASN > 0 {
-		if err := h.db.DeleteLiveRoute(update.prefix, update.originASN, update.peer); err != nil {
-			h.logger.Error("Failed to delete live route",
-				"prefix", update.prefix,
-				"origin_asn", update.originASN,
-				"peer", update.peer,
-				"error", err,
-			)
-		}
-	} else {
-		// If no origin ASN, just delete all routes for this prefix from this peer
-		if err := h.db.DeleteLiveRoute(update.prefix, 0, update.peer); err != nil {
-			h.logger.Error("Failed to delete live route (no origin ASN)",
-				"prefix", update.prefix,
-				"peer", update.peer,
-				"error", err,
-			)
-		}
-	}
-}
-
-// Stop gracefully stops the handler and flushes remaining batches
-func (h *PrefixHandler) Stop() {
-	close(h.stopCh)
-	h.wg.Wait()
-}

internal/routewatch/prefixhandler_table_internal_test.go

@@ -0,0 +1,535 @@
+package routewatch
+
+import (
+	"bytes"
+	"encoding/json"
+	"errors"
+	"log/slog"
+	"net/netip"
+	"reflect"
+	"slices"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/ristypes"
+	"git.eeqj.de/sneak/routewatch/internal/routetable"
+)
+
+// More peers and prefixes for the tests of the routing table in memory.
+const (
+	secondPeerIP = "2001:db8:ffff::2"
+	thirdPeerIP  = "2001:db8:ffff::3"
+	fourthPeerIP = "192.0.2.4"
+	fifthPeerIP  = "192.0.2.5"
+
+	otherV4Prefix = "198.51.100.0/24"
+	setV4Prefix   = "203.0.113.0/24"
+
+	// firstManyPrefix is the first prefix announceManyPrefixes announces.
+	firstManyPrefix = "10.0.0.0/32"
+)
+
+// tableWriteTimeout is how long a test waits for the flush timer to write the
+// routing table's changes: a few ticks of it.
+const tableWriteTimeout = 5 * prefixBatchTimeout
+
+// tableCase is one test of the routing table in memory: the messages handed to
+// the prefix handler, and for each prefix the peers that carry it with each
+// origin AS after them.
+type tableCase struct {
+	name     string
+	messages []*ristypes.RISMessage
+	want     map[string]map[int][]string
+}
+
+// TestPrefixHandlerAppliesUpdatesToTheRoutingTable hands the prefix handler
+// announcements, withdrawals and a peer's session going down.
+func TestPrefixHandlerAppliesUpdatesToTheRoutingTable(t *testing.T) {
+	t.Parallel()
+
+	announceAndWithdraw := peerAnnouncement(testPeerIP, testV4Prefix)
+	announceAndWithdraw.Withdrawals = []string{testV4Prefix}
+
+	checkTableCases(t, []tableCase{
+		{
+			name:     "an announcement",
+			messages: []*ristypes.RISMessage{peerAnnouncement(testPeerIP, testV4Prefix)},
+			want:     map[string]map[int][]string{testV4Prefix: {testASNB: {testPeerIP}}},
+		},
+		{
+			name: "a withdrawal",
+			messages: []*ristypes.RISMessage{
+				peerAnnouncement(testPeerIP, testV4Prefix),
+				peerWithdrawal(testPeerIP, testV4Prefix),
+			},
+			want: map[string]map[int][]string{testV4Prefix: {}},
+		},
+		{
+			// RFC 4271 section 4.3: the announcement counts, not the withdrawal.
+			name:     "a message that announces and withdraws the prefix",
+			messages: []*ristypes.RISMessage{announceAndWithdraw},
+			want:     map[string]map[int][]string{testV4Prefix: {testASNB: {testPeerIP}}},
+		},
+		{
+			name: "a peer's session going down",
+			messages: []*ristypes.RISMessage{
+				peerAnnouncement(testPeerIP, testV4Prefix, testV6Prefix),
+				peerAnnouncement(secondPeerIP, testV4Prefix),
+				sessionDown(),
+			},
+			want: map[string]map[int][]string{
+				testV4Prefix: {testASNB: {secondPeerIP}},
+				testV6Prefix: {},
+			},
+		},
+	})
+}
+
+// TestPrefixHandlerAppliesAPathWithoutAnOriginToTheRoutingTable hands the
+// prefix handler announcements whose AS path gives no origin. One whose path
+// holds AS 0 removes the peer's route, as a withdrawal does; one whose path ends
+// in an AS set of several members is kept with origin 0.
+func TestPrefixHandlerAppliesAPathWithoutAnOriginToTheRoutingTable(t *testing.T) {
+	t.Parallel()
+
+	announceWithPath := func(path ristypes.ASPath) *ristypes.RISMessage {
+		msg := peerAnnouncement(testPeerIP, testV4Prefix)
+		msg.Path = path
+
+		return msg
+	}
+
+	checkTableCases(t, []tableCase{
+		{
+			name: "a path whose last AS is 0",
+			messages: []*ristypes.RISMessage{
+				peerAnnouncement(testPeerIP, testV4Prefix),
+				announceWithPath(ristypes.ASPath{Sequence: []int{testASNA, 0}}),
+			},
+			want: map[string]map[int][]string{testV4Prefix: {}},
+		},
+		{
+			name: "a path with AS 0 before its end",
+			messages: []*ristypes.RISMessage{
+				peerAnnouncement(testPeerIP, testV4Prefix),
+				announceWithPath(decodePath(t, testASNA, 0, testASNB)),
+			},
+			want: map[string]map[int][]string{testV4Prefix: {}},
+		},
+		{
+			name: "a path that ends in an AS set",
+			messages: []*ristypes.RISMessage{
+				announceWithPath(decodePath(t, testASNA, []int{testASNC, testASND})),
+			},
+			want: map[string]map[int][]string{testV4Prefix: {0: {testPeerIP}}},
+		},
+	})
+}
+
+// checkTableCases hands the messages of each case to a new prefix handler and
+// checks its routing table right after, with no flush.
+func checkTableCases(t *testing.T, tests []tableCase) {
+	t.Helper()
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			t.Parallel()
+
+			h := prefixHandlerWithoutTimer(newMockStore())
+			for _, msg := range tt.messages {
+				h.HandleMessage(msg)
+			}
+
+			for prefix, want := range tt.want {
+				got := tableOrigins(h.table, prefix)
+				if !reflect.DeepEqual(got, want) {
+					t.Errorf("peers of %s by origin AS = %v, want %v",
+						prefix, got, want)
+				}
+			}
+		})
+	}
+}
+
+// tableWritesStore is a real database that hands each write of the routing
+// table's changes to written once it is made.
+type tableWritesStore struct {
+	*database.Database
+
+	written chan []routetable.Change
+}
+
+func (s tableWritesStore) WriteRouteChanges(
+	changes []routetable.Change, newPeers map[string]uint,
+) error {
+	err := s.Database.WriteRouteChanges(changes, newPeers)
+	s.written <- changes
+
+	return err
+}
+
+// TestPrefixHandlerWritesTheTableChangesEachSecond hands a running prefix
+// handler an announcement and then its withdrawal, each far short of a full
+// batch. The flush timer must write the change of each to the database.
+func TestPrefixHandlerWritesTheTableChangesEachSecond(t *testing.T) {
+	t.Parallel()
+
+	// The two writes the test waits for, and room for one more.
+	const writesBuffered = 3
+
+	store := tableWritesStore{
+		Database: newTestDatabase(t),
+		written:  make(chan []routetable.Change, writesBuffered),
+	}
+	h := NewPrefixHandler(store, logger.New(), routetable.New(routetable.Params{}))
+
+	defer h.Stop(t.Context())
+
+	prefix := netip.MustParsePrefix(testV4Prefix)
+
+	for _, step := range []struct {
+		msg         *ristypes.RISMessage
+		wantRemoved bool
+	}{
+		{msg: peerAnnouncement(testPeerIP, testV4Prefix)},
+		{msg: peerWithdrawal(testPeerIP, testV4Prefix), wantRemoved: true},
+	} {
+		h.HandleMessage(step.msg)
+
+		select {
+		case changes := <-store.written:
+			if len(changes) != 1 || changes[0].Prefix != prefix ||
+				changes[0].OriginASN != testASNB ||
+				changes[0].Removed != step.wantRemoved {
+				t.Fatalf("wrote %+v, want the pair of %s and AS%d with "+
+					"removed %t", changes, prefix, testASNB, step.wantRemoved)
+			}
+		case <-time.After(tableWriteTimeout):
+			t.Fatalf("the routing table's change was not written within %s",
+				tableWriteTimeout)
+		}
+	}
+}
+
+// errTestTableWrite is the error a test store's failing write of the routing
+// table's changes returns.
+var errTestTableWrite = errors.New("test: the write of the table's changes failed")
+
+// firstTableWriteFailsStore is a real database whose first write of the
+// routing table's changes fails, writing nothing. writes counts the writes
+// asked for.
+type firstTableWriteFailsStore struct {
+	*database.Database
+
+	writes int
+}
+
+func (s *firstTableWriteFailsStore) WriteRouteChanges(
+	changes []routetable.Change, newPeers map[string]uint,
+) error {
+	s.writes++
+	if s.writes == 1 {
+		return errTestTableWrite
+	}
+
+	return s.Database.WriteRouteChanges(changes, newPeers)
+}
+
+// TestPrefixHandlerWritesAFailedTableWriteAgain runs two flushes into a
+// database whose first write of the routing table's changes fails. Before the
+// first, two new peers announce a prefix each; before the second, one of them
+// announces its prefix again with another path to the same origin AS. The
+// second write must hold the changes and new peers of the first, with the later
+// path, so the database holds the table that is in memory.
+func TestPrefixHandlerWritesAFailedTableWriteAgain(t *testing.T) {
+	t.Parallel()
+
+	db := newTestDatabase(t)
+	store := &firstTableWriteFailsStore{Database: db}
+	h := prefixHandlerWithoutTimer(store)
+
+	first := peerAnnouncement(testPeerIP, testV4Prefix)
+	h.HandleMessage(first)
+	h.HandleMessage(peerAnnouncement(secondPeerIP, otherV4Prefix))
+	h.flushBatch()
+
+	laterPath := []int{testASNC, testASNB}
+	later := peerAnnouncement(testPeerIP, testV4Prefix)
+	later.Path = ristypes.ASPath{Sequence: laterPath}
+	later.ParsedTimestamp = first.ParsedTimestamp.Add(secondUpdateAfter)
+	h.HandleMessage(later)
+	h.flushBatch()
+
+	if store.writes != 2 {
+		t.Fatalf("the store was asked for %d writes of the table's changes, "+
+			"want 2", store.writes)
+	}
+
+	read, err := db.ReadRouteTable(t.Context())
+	if err != nil {
+		t.Fatalf("ReadRouteTable: %v", err)
+	}
+
+	checkSameTable(t, read, h.table, testV4Prefix, otherV4Prefix)
+
+	origins := read.Origins(netip.MustParsePrefix(testV4Prefix))
+	if len(origins) != 1 || len(origins[0].Routes) != 1 ||
+		!slices.Equal(origins[0].Routes[0].ASPath, laterPath) {
+		t.Errorf("Origins(%s) = %+v, want one route with the path %v",
+			testV4Prefix, origins, laterPath)
+	}
+}
+
+// tableWriteSizesStore is a real database that records how many changes and
+// how many new peers each write of the routing table's changes holds.
+type tableWriteSizesStore struct {
+	*database.Database
+
+	changes  []int
+	newPeers []int
+}
+
+func (s *tableWriteSizesStore) WriteRouteChanges(
+	changes []routetable.Change, newPeers map[string]uint,
+) error {
+	s.changes = append(s.changes, len(changes))
+	s.newPeers = append(s.newPeers, len(newPeers))
+
+	return s.Database.WriteRouteChanges(changes, newPeers)
+}
+
+// TestPrefixHandlerWritesManyTableChangesOverSeveralFlushes has one new peer
+// change one and a half times as many prefix and origin AS pairs as one write
+// takes, then flushes three times. The first write must hold as many pairs as
+// one write takes, and the peer. The second must hold the other pairs, and the
+// third flush must write nothing. The database must then hold the table in
+// memory.
+func TestPrefixHandlerWritesManyTableChangesOverSeveralFlushes(t *testing.T) {
+	t.Parallel()
+
+	const (
+		pairs   = maxTableChangesPerWrite * 3 / 2
+		flushes = 3
+	)
+
+	db := newTestDatabase(t)
+	store := &tableWriteSizesStore{Database: db}
+	h := prefixHandlerWithoutTimer(store)
+	announceManyPrefixes(h.table, pairs)
+
+	for range flushes {
+		h.flushBatch()
+	}
+
+	wantChanges := []int{maxTableChangesPerWrite, pairs - maxTableChangesPerWrite}
+	wantNewPeers := []int{1, 0}
+
+	if !slices.Equal(store.changes, wantChanges) ||
+		!slices.Equal(store.newPeers, wantNewPeers) {
+		t.Errorf("the writes held %v pairs and %v new peers, want %v and %v",
+			store.changes, store.newPeers, wantChanges, wantNewPeers)
+	}
+
+	read, err := db.ReadRouteTable(t.Context())
+	if err != nil {
+		t.Fatalf("ReadRouteTable: %v", err)
+	}
+
+	checkSameTable(t, read, h.table, firstManyPrefix)
+}
+
+// TestPrefixHandlerStopLogsTheTableChangesLeftUnwritten stops a prefix handler,
+// with no deadline, whose routing table holds more changed pairs than two
+// writes take. Its last flush writes one write's worth, so Stop must log the
+// pairs that no write took.
+func TestPrefixHandlerStopLogsTheTableChangesLeftUnwritten(t *testing.T) {
+	t.Parallel()
+
+	// More than two writes take, so that pairs are left even when the flush
+	// timer writes once before Stop.
+	const pairs = 2*maxTableChangesPerWrite + 1
+
+	var logs bytes.Buffer
+
+	store := &tableWriteSizesStore{Database: newTestDatabase(t)}
+	table := routetable.New(routetable.Params{})
+	announceManyPrefixes(table, pairs)
+
+	h := NewPrefixHandler(store, newErrorLogger(&logs), table)
+	h.Stop(t.Context())
+
+	written := 0
+	for _, n := range store.changes {
+		written += n
+	}
+
+	if logged := lastLogged(t, &logs); logged.TableChanges != pairs-written {
+		t.Errorf("Stop logged %d table changes not written, want %d",
+			logged.TableChanges, pairs-written)
+	}
+}
+
+// readLog holds what readRouteTable logs.
+//
+//nolint:tagliatelle // field names are the keys readRouteTable logs
+type readLog struct {
+	Msg        string `json:"msg"`
+	Pairs      int    `json:"pairs"`
+	Peers      int    `json:"peers"`
+	DurationMs *int64 `json:"duration_ms"`
+}
+
+// TestARestartReadsTheRoutingTableBack runs announcements, a withdrawal and a
+// peer's session going down through a prefix handler into a database, then
+// reads the table back as startup does. The table read must answer as the one
+// in memory, and the log must give its pairs and peers. A new handler on the
+// table read then takes later updates and writes them as it stops: the
+// database must then hold what the first table holds after the same updates.
+func TestARestartReadsTheRoutingTableBack(t *testing.T) {
+	t.Parallel()
+
+	prefixes := []string{testV4Prefix, testV6Prefix, otherV4Prefix, setV4Prefix}
+
+	db := newTestDatabase(t)
+	h := prefixHandlerWithoutTimer(db)
+
+	viaOtherOrigin := peerAnnouncement(thirdPeerIP, testV4Prefix)
+	viaOtherOrigin.Path = ristypes.ASPath{Sequence: []int{testASNA, testASNC}}
+	endsInSet := peerAnnouncement(secondPeerIP, setV4Prefix)
+	endsInSet.Path = decodePath(t, testASNA, []int{testASNC, testASND})
+
+	for _, msg := range []*ristypes.RISMessage{
+		peerAnnouncement(testPeerIP, testV4Prefix, testV6Prefix),
+		peerAnnouncement(secondPeerIP, testV4Prefix),
+		viaOtherOrigin,
+		endsInSet,
+		peerAnnouncement(thirdPeerIP, otherV4Prefix),
+		peerWithdrawal(thirdPeerIP, otherV4Prefix),
+		peerAnnouncement(fourthPeerIP, otherV4Prefix),
+		sessionDownOf(fourthPeerIP),
+	} {
+		h.HandleMessage(msg)
+	}
+
+	h.flushBatch()
+
+	var logs bytes.Buffer
+
+	rw := &RouteWatch{
+		db:     db,
+		logger: &logger.Logger{Logger: slog.New(slog.NewJSONHandler(&logs, nil))},
+	}
+
+	table, err := rw.readRouteTable(t.Context())
+	if err != nil {
+		t.Fatalf("readRouteTable: %v", err)
+	}
+
+	checkSameTable(t, table, h.table, prefixes...)
+
+	// testV4Prefix with two origin ASes, testV6Prefix and setV4Prefix; four
+	// peers, the last of which carries nothing.
+	const wantPairs, wantPeers = 4, 4
+
+	var logged readLog
+
+	err = json.Unmarshal(bytes.TrimSpace(logs.Bytes()), &logged)
+	if err != nil || logged.Pairs != wantPairs || logged.Peers != wantPeers ||
+		logged.DurationMs == nil {
+		t.Errorf("readRouteTable logged %q, want %d pairs, %d peers and the "+
+			"duration", logs.String(), wantPairs, wantPeers)
+	}
+
+	restarted := NewPrefixHandler(db, logger.New(), table)
+
+	for _, msg := range []*ristypes.RISMessage{
+		peerWithdrawal(testPeerIP, testV6Prefix),
+		sessionDownOf(secondPeerIP),
+		peerAnnouncement(fourthPeerIP, otherV4Prefix),
+		peerAnnouncement(fifthPeerIP, setV4Prefix),
+	} {
+		// The first handler only applies the updates to its table.
+		h.HandleMessage(msg)
+		restarted.HandleMessage(msg)
+	}
+
+	restarted.Stop(t.Context())
+
+	read, err := db.ReadRouteTable(t.Context())
+	if err != nil {
+		t.Fatalf("ReadRouteTable after the restart: %v", err)
+	}
+
+	checkSameTable(t, read, h.table, prefixes...)
+}
+
+// peerWithdrawal returns an UPDATE message in which peer withdraws prefixes.
+func peerWithdrawal(peer string, prefixes ...string) *ristypes.RISMessage {
+	return &ristypes.RISMessage{
+		Type:            ristypes.MessageTypeUpdate,
+		Peer:            peer,
+		ParsedTimestamp: time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC),
+		Withdrawals:     prefixes,
+	}
+}
+
+// announceManyPrefixes has testPeerIP announce n prefixes in table, with the
+// path of peerAnnouncement: the /32 of each address from firstManyPrefix on.
+func announceManyPrefixes(table *routetable.Table, n int) {
+	addr := netip.MustParsePrefix(firstManyPrefix).Addr()
+	route := routetable.Route{
+		Peer:   testPeerIP,
+		ASPath: []int{testASNA, testASNB},
+		Time:   time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC),
+	}
+
+	for range n {
+		table.Announce(netip.PrefixFrom(addr, addr.BitLen()), testASNB, route)
+		addr = addr.Next()
+	}
+}
+
+// sessionDownOf returns the STATE message that says the BGP session of peer
+// went down.
+func sessionDownOf(peer string) *ristypes.RISMessage {
+	return &ristypes.RISMessage{
+		Type:  ristypes.MessageTypeState,
+		Peer:  peer,
+		State: ristypes.StateDown,
+	}
+}
+
+// tableOrigins returns, for each origin AS of prefix in table, the peers that
+// carry the prefix with it.
+func tableOrigins(table *routetable.Table, prefix string) map[int][]string {
+	origins := make(map[int][]string)
+	for _, o := range table.Origins(netip.MustParsePrefix(prefix)) {
+		origins[o.ASN] = o.Peers
+	}
+
+	return origins
+}
+
+// checkSameTable checks that got answers as want does for each of prefixes, and
+// that the two tables have the same counts.
+func checkSameTable(
+	t *testing.T, got, want *routetable.Table, prefixes ...string,
+) {
+	t.Helper()
+
+	for _, text := range prefixes {
+		prefix := netip.MustParsePrefix(text)
+
+		gotOrigins, wantOrigins := got.Origins(prefix), want.Origins(prefix)
+		if !reflect.DeepEqual(gotOrigins, wantOrigins) {
+			t.Errorf("Origins(%s) = %+v, want %+v", text, gotOrigins, wantOrigins)
+		}
+	}
+
+	gotCounts, wantCounts := got.Counts(), want.Counts()
+	if !reflect.DeepEqual(gotCounts, wantCounts) {
+		t.Errorf("Counts() = %+v, want %+v", gotCounts, wantCounts)
+	}
+}

internal/routewatch/prefixhandler_test.go

@@ -1,74 +0,0 @@
-package routewatch
-
-import (
-	"errors"
-	"testing"
-	"time"
-
-	"git.eeqj.de/sneak/routewatch/internal/config"
-	"git.eeqj.de/sneak/routewatch/internal/database"
-	"git.eeqj.de/sneak/routewatch/internal/logger"
-	"git.eeqj.de/sneak/routewatch/internal/ristypes"
-)
-
-const testPeerIP = "2001:db8:ffff::1"
-
-// TestPrefixHandlerStoresPrefixesTheIPLookupFinds runs announcements and
-// withdrawals through the prefix handler into a real database and looks the
-// addresses up. The prefixes are written the way the feed sends them: IPv6
-// announcements compressed, IPv6 withdrawals uncompressed. The withdrawal must
-// remove the route the announcement stored, and the IP lookup must find the
-// stored prefix.
-func TestPrefixHandlerStoresPrefixesTheIPLookupFinds(t *testing.T) {
-	db, err := database.New(&config.Config{StateDir: t.TempDir()}, logger.New())
-	if err != nil {
-		t.Fatalf("failed to create database: %v", err)
-	}
-	defer func() { _ = db.Close() }()
-
-	// Built without NewPrefixHandler's flush timer, so the test flushes itself.
-	h := &PrefixHandler{db: db, logger: logger.New()}
-	flush := func() {
-		h.mu.Lock()
-		defer h.mu.Unlock()
-		h.flushBatchLocked()
-	}
-
-	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
-	h.HandleMessage(&ristypes.RISMessage{
-		Peer:            testPeerIP,
-		Path:            ristypes.ASPath{testASNA, testASNB},
-		ParsedTimestamp: ts,
-		Announcements: []ristypes.RISAnnouncement{{
-			NextHop:  testPeerIP,
-			Prefixes: []string{"2001:db8:1::/48", "192.0.2.0/24"},
-		}},
-	})
-	flush()
-
-	for ip, want := range map[string]string{
-		"2001:db8:1::1": "2001:db8:1::/48",
-		"192.0.2.1":     "192.0.2.0/24",
-	} {
-		info, err := db.GetIPInfo(ip)
-		if err != nil {
-			t.Fatalf("GetIPInfo(%s) after announcement: %v", ip, err)
-		}
-		if info.Netblock != want || info.ASN != testASNB {
-			t.Errorf("GetIPInfo(%s) = %s AS%d, want %s AS%d", ip, info.Netblock, info.ASN, want, testASNB)
-		}
-	}
-
-	h.HandleMessage(&ristypes.RISMessage{
-		Peer:            testPeerIP,
-		ParsedTimestamp: ts.Add(time.Minute),
-		Withdrawals:     []string{"2001:db8:1:0:0:0:0:0/48", "192.0.2.0/24"},
-	})
-	flush()
-
-	for _, ip := range []string{"2001:db8:1::1", "192.0.2.1"} {
-		if info, err := db.GetIPInfo(ip); !errors.Is(err, database.ErrNoRoute) {
-			t.Errorf("GetIPInfo(%s) after withdrawal = %+v, %v; want ErrNoRoute", ip, info, err)
-		}
-	}
-}

internal/server/handlers_test.go

@@ -1,177 +0,0 @@
-package server
-
-import (
-	"context"
-	"encoding/json"
-	"net/http"
-	"net/http/httptest"
-	"runtime"
-	"slices"
-	"testing"
-	"time"
-
-	"git.eeqj.de/sneak/routewatch/internal/config"
-	"git.eeqj.de/sneak/routewatch/internal/database"
-	"git.eeqj.de/sneak/routewatch/internal/logger"
-	"git.eeqj.de/sneak/routewatch/internal/metrics"
-	"git.eeqj.de/sneak/routewatch/internal/streamer"
-	"github.com/google/uuid"
-)
-
-// blockingStatsDB embeds database.Store (left nil) and overrides only
-// GetStatsContext, which blocks until release is closed. The stats handlers
-// call it in a goroutine; every other Store method is unused on the timeout
-// path and would panic if called.
-type blockingStatsDB struct {
-	database.Store
-	release chan struct{}
-}
-
-func (d blockingStatsDB) GetStatsContext(_ context.Context) (database.Stats, error) {
-	<-d.release
-
-	return database.Stats{}, nil
-}
-
-// TestStatsHandlersDoNotLeakOnTimeout drives each stats handler repeatedly with
-// a request whose context times out before the database responds, then releases
-// the blocked queries and asserts the goroutine count returns to its starting
-// value. Before the fix the per-request goroutine sent on an unbuffered channel
-// that nothing received once the timeout won, so it blocked forever and every
-// poll leaked one goroutine.
-func TestStatsHandlersDoNotLeakOnTimeout(t *testing.T) {
-	release := make(chan struct{})
-	db := blockingStatsDB{release: release}
-	s := New(db, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})
-
-	handlers := map[string]http.HandlerFunc{
-		"status.json": s.handleStatusJSON(),
-		"stats":       s.handleStats(),
-	}
-
-	baseline := settledGoroutineCount()
-
-	const (
-		iterations     = 20
-		requestTimeout = 50 * time.Millisecond
-	)
-	for _, handler := range handlers {
-		for range iterations {
-			ctx, cancel := context.WithTimeout(context.Background(), requestTimeout)
-			req := httptest.NewRequest(http.MethodGet, "/", nil).WithContext(ctx)
-			handler(httptest.NewRecorder(), req)
-			cancel()
-		}
-	}
-
-	// Let the blocked queries finish; with buffered channels each goroutine's
-	// send now succeeds and the goroutine exits.
-	close(release)
-
-	if !waitForGoroutines(baseline) {
-		t.Fatalf("goroutines did not return to baseline %d, got %d",
-			baseline, runtime.NumGoroutine())
-	}
-}
-
-// TestStatsHandlersAnswerFromMemory checks that both stats handlers answer 200
-// with the live route counts and the prefix distribution while the database is
-// closed, so that any query would fail: the request path reads them from
-// memory. The prefix distribution query it used to run read every live route
-// and, on a large database, took the whole 4-second deadline, so
-// /api/v1/stats answered 500 (https://git.eeqj.de/sneak/routewatch/issues/30).
-// The oldest and newest route times still come from one-row lookups at the ends
-// of an index; with the database closed they are left out of the answer.
-func TestStatsHandlersAnswerFromMemory(t *testing.T) {
-	db, err := database.New(&config.Config{StateDir: t.TempDir()}, logger.New())
-	if err != nil {
-		t.Fatalf("database.New: %v", err)
-	}
-
-	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
-	if err := db.UpsertLiveRouteBatch([]*database.LiveRoute{
-		{
-			ID: uuid.New(), Prefix: "198.51.100.0/24", MaskLength: 24, IPVersion: 4,
-			OriginASN: 64500, PeerIP: "192.0.2.1", ASPath: []int{64500}, NextHop: "192.0.2.1",
-			LastUpdated: ts,
-		},
-		{
-			ID: uuid.New(), Prefix: "2001:db8::/32", MaskLength: 32, IPVersion: 6,
-			OriginASN: 64501, PeerIP: "2001:db8::1", ASPath: []int{64501}, NextHop: "2001:db8::1",
-			LastUpdated: ts,
-		},
-	}); err != nil {
-		t.Fatalf("UpsertLiveRouteBatch: %v", err)
-	}
-	if err := db.Close(); err != nil {
-		t.Fatalf("Close: %v", err)
-	}
-
-	s := New(db, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})
-	handlers := map[string]http.HandlerFunc{
-		"status.json": s.handleStatusJSON(),
-		"stats":       s.handleStats(),
-	}
-
-	for name, handler := range handlers {
-		rec := httptest.NewRecorder()
-		handler(rec, httptest.NewRequest(http.MethodGet, "/", nil))
-		if rec.Code != http.StatusOK {
-			t.Errorf("%s: status %d, want %d; body %s", name, rec.Code, http.StatusOK, rec.Body)
-
-			continue
-		}
-
-		var body struct {
-			Data struct {
-				IPv4Routes             int                           `json:"ipv4_routes"`
-				IPv6Routes             int                           `json:"ipv6_routes"`
-				IPv4PrefixDistribution []database.PrefixDistribution `json:"ipv4_prefix_distribution"`
-				IPv6PrefixDistribution []database.PrefixDistribution `json:"ipv6_prefix_distribution"`
-			} `json:"data"`
-		}
-		if err := json.Unmarshal(rec.Body.Bytes(), &body); err != nil {
-			t.Fatalf("%s: decoding the answer: %v", name, err)
-		}
-
-		if body.Data.IPv4Routes != 1 || body.Data.IPv6Routes != 1 {
-			t.Errorf("%s: routes = (v4 %d, v6 %d), want (1, 1)", name, body.Data.IPv4Routes, body.Data.IPv6Routes)
-		}
-		wantV4 := []database.PrefixDistribution{{MaskLength: 24, Count: 1}}
-		if !slices.Equal(body.Data.IPv4PrefixDistribution, wantV4) {
-			t.Errorf("%s: IPv4 distribution = %v, want %v", name, body.Data.IPv4PrefixDistribution, wantV4)
-		}
-		wantV6 := []database.PrefixDistribution{{MaskLength: 32, Count: 1}}
-		if !slices.Equal(body.Data.IPv6PrefixDistribution, wantV6) {
-			t.Errorf("%s: IPv6 distribution = %v, want %v", name, body.Data.IPv6PrefixDistribution, wantV6)
-		}
-	}
-}
-
-// settledGoroutineCount lets transient goroutines finish, then reports the
-// current count.
-func settledGoroutineCount() int {
-	prev := runtime.NumGoroutine()
-	for range 20 {
-		time.Sleep(10 * time.Millisecond)
-		cur := runtime.NumGoroutine()
-		if cur == prev {
-			return cur
-		}
-		prev = cur
-	}
-
-	return prev
-}
-
-// waitForGoroutines waits until the goroutine count drops to target or below.
-func waitForGoroutines(target int) bool {
-	for range 100 {
-		if runtime.NumGoroutine() <= target {
-			return true
-		}
-		time.Sleep(10 * time.Millisecond)
-	}
-
-	return false
-}

internal/server/middleware.go

@@ -3,17 +3,25 @@ package server
 import (
 	"bytes"
 	"context"
+	"crypto/rand"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"log/slog"
 	"net/http"
+	"net/netip"
+	"slices"
+	"strings"
 	"sync"
 	"time"
+
+	"github.com/go-chi/chi/v5"
 )
 
 // responseWriter wraps http.ResponseWriter to capture the response
 type responseWriter struct {
 	http.ResponseWriter
+
 	body       *bytes.Buffer
 	statusCode int
 	written    bool
@@ -86,8 +94,10 @@ func JSONResponseMiddleware(next http.Handler) http.Handler {
 		}
 
 		// Parse the original response
-		var originalResponse map[string]interface{}
+		var originalResponse map[string]any
-		if err := json.Unmarshal(rw.body.Bytes(), &originalResponse); err != nil {
+
+		err := json.Unmarshal(rw.body.Bytes(), &originalResponse)
+		if err != nil {
 			// If we can't parse it, just write original
 			w.WriteHeader(rw.statusCode)
 			_, _ = w.Write(rw.body.Bytes())
@@ -96,7 +106,7 @@ func JSONResponseMiddleware(next http.Handler) http.Handler {
 		}
 
 		// Add @meta field
-		originalResponse["@meta"] = map[string]interface{}{
+		originalResponse["@meta"] = map[string]any{
 			"time_zone":      "UTC",
 			"api_version":    1,
 			"execution_time": fmt.Sprintf("%d ms", execTime.Milliseconds()),
@@ -109,142 +119,33 @@ func JSONResponseMiddleware(next http.Handler) http.Handler {
 	})
 }
 
-// timeoutWriter wraps ResponseWriter to prevent concurrent writes after timeout
+// requestTimeoutMessage is the error message sent with 504 Gateway Timeout.
-type timeoutWriter struct {
+const requestTimeoutMessage = "Request timeout"
-	http.ResponseWriter
-	mu      sync.Mutex
-	written bool
-	header  http.Header // cached header to prevent concurrent access
-}
 
-func (tw *timeoutWriter) Write(b []byte) (int, error) {
+// internalErrorMessage is the error message sent with 500 Internal Server
-	tw.mu.Lock()
+// Error, whatever went wrong; the handler logs the detail.
-	defer tw.mu.Unlock()
+const internalErrorMessage = "Internal server error"
 
-	if tw.written {
+// deadlinePassed reports whether ctx ended because its deadline passed.
-		return 0, nil // Discard writes after timeout
-	}
-
-	return tw.ResponseWriter.Write(b)
-}
-
-func (tw *timeoutWriter) WriteHeader(statusCode int) {
-	tw.mu.Lock()
-	defer tw.mu.Unlock()
-
-	if tw.written {
-		return // Discard writes after timeout
-	}
-
-	tw.ResponseWriter.WriteHeader(statusCode)
-}
-
-func (tw *timeoutWriter) Header() http.Header {
-	tw.mu.Lock()
-	defer tw.mu.Unlock()
-
-	if tw.written {
-		// Return a copy to prevent modifications after timeout
-		if tw.header == nil {
-			tw.header = make(http.Header)
-		}
-
-		return tw.header
-	}
-
-	return tw.ResponseWriter.Header()
-}
-
-func (tw *timeoutWriter) markWritten() {
-	tw.mu.Lock()
-	defer tw.mu.Unlock()
-	tw.written = true
-}
-
-// TimeoutMiddleware creates an HTTP middleware that enforces a request timeout.
-// If the handler does not complete within the specified duration, the middleware
-// returns a JSON error response with HTTP status 408 (Request Timeout).
 //
-// The timeout parameter specifies the maximum duration allowed for request processing.
+// A handler calls it right after each database read or host name lookup,
-// The returned middleware handles panics from the wrapped handler by re-panicking
+// before it looks at what came back, error or not, and answers 504 if it
-// after cleanup, and prevents concurrent writes to the response after timeout occurs.
+// reports true. A read cut off by the time limit can return an error, or part
-func TimeoutMiddleware(timeout time.Duration, logger *slog.Logger) func(http.Handler) http.Handler {
+// of its rows or none of them without one, so only this check tells a cut-off
-	return func(next http.Handler) http.Handler {
+// read from a complete one.
-		return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+func deadlinePassed(ctx context.Context) bool {
-			startTime := time.Now()
+	return errors.Is(ctx.Err(), context.DeadlineExceeded)
-
-			ctx, cancel := context.WithTimeout(r.Context(), timeout)
-			defer cancel()
-
-			tw := &timeoutWriter{
-				ResponseWriter: w,
-				header:         make(http.Header),
-			}
-
-			done := make(chan struct{})
-			panicChan := make(chan interface{}, 1)
-
-			go func() {
-				defer func() {
-					if p := recover(); p != nil {
-						panicChan <- p
-					}
-				}()
-
-				next.ServeHTTP(tw, r.WithContext(ctx))
-				close(done)
-			}()
-
-			select {
-			case p := <-panicChan:
-				panic(p)
-			case <-done:
-				return
-			case <-ctx.Done():
-				tw.markWritten() // Prevent the handler from writing after timeout
-				execTime := time.Since(startTime)
-
-				// Log the timeout as a warning
-				logger.Warn("Request timeout",
-					"method", r.Method,
-					"path", r.URL.Path,
-					"duration_ms", execTime.Milliseconds(),
-					"remote_addr", r.RemoteAddr,
-				)
-
-				// Write directly to the underlying writer since we've marked tw as written
-				// This is safe because markWritten() prevents the handler from writing
-				tw.mu.Lock()
-				w.Header().Set("Content-Type", "application/json")
-				w.WriteHeader(http.StatusRequestTimeout)
-				tw.mu.Unlock()
-
-				response := map[string]interface{}{
-					"status": "error",
-					"error": map[string]interface{}{
-						"msg":  "Request timeout",
-						"code": http.StatusRequestTimeout,
-					},
-					"@meta": map[string]interface{}{
-						"time_zone":      "UTC",
-						"api_version":    1,
-						"execution_time": fmt.Sprintf("%d ms", execTime.Milliseconds()),
-					},
-				}
-
-				_ = json.NewEncoder(w).Encode(response)
-			}
-		})
-	}
 }
 
 // JSONValidationMiddleware is an HTTP middleware that validates JSON API responses.
 // It ensures that responses with Content-Type "application/json" contain valid JSON.
 //
 // If a response is not valid JSON or is empty when JSON is expected, the middleware
-// returns a properly formatted JSON error response. For timeout errors (status 408),
+// returns a properly formatted JSON error response. If the request's time limit
-// the error message will be "Request timeout". For other errors, it returns
+// has passed, it answers 504 with the error message "Request timeout". A 4xx
-// "Internal server error" with status 500 if the original status was 200.
+// keeps its status, with the status's standard text as the message. For other
+// errors, it returns "Internal server error", with status 500 if the original
+// status was 200.
 //
 // Non-JSON responses are passed through unchanged.
 func JSONValidationMiddleware(next http.Handler) http.Handler {
@@ -273,8 +174,10 @@ func JSONValidationMiddleware(next http.Handler) http.Handler {
 
 		// For JSON responses, validate the JSON
 		if rw.body.Len() > 0 {
-			var testParse interface{}
+			var testParse any
-			if err := json.Unmarshal(rw.body.Bytes(), &testParse); err == nil {
+
+			err := json.Unmarshal(rw.body.Bytes(), &testParse)
+			if err == nil {
 				// Valid JSON, write it out
 				w.Header().Set("Content-Type", "application/json")
 				w.WriteHeader(rw.statusCode)
@@ -293,18 +196,28 @@ func JSONValidationMiddleware(next http.Handler) http.Handler {
 		if statusCode == http.StatusOK {
 			statusCode = http.StatusInternalServerError
 		}
-		w.WriteHeader(statusCode)
 
-		errorMsg := "Internal server error"
+		if deadlinePassed(r.Context()) {
-		if statusCode == http.StatusRequestTimeout {
+			statusCode = http.StatusGatewayTimeout
-			errorMsg = "Request timeout"
 		}
 
-		response := map[string]interface{}{
+		w.WriteHeader(statusCode)
-			"status": "error",
+
-			"error": map[string]interface{}{
+		errorMsg := internalErrorMessage
-				"msg":  errorMsg,
+
-				"code": statusCode,
+		switch {
+		case statusCode == http.StatusGatewayTimeout:
+			errorMsg = requestTimeoutMessage
+		case statusCode >= http.StatusBadRequest &&
+			statusCode < http.StatusInternalServerError:
+			errorMsg = http.StatusText(statusCode)
+		}
+
+		response := map[string]any{
+			jsonKeyStatus: jsonStatusError,
+			jsonKeyError: map[string]any{
+				jsonKeyMsg:  errorMsg,
+				jsonKeyCode: statusCode,
 			},
 		}
 		_ = json.NewEncoder(w).Encode(response)
@@ -314,6 +227,7 @@ func JSONValidationMiddleware(next http.Handler) http.Handler {
 // statusWriter wraps http.ResponseWriter to capture the status code
 type statusWriter struct {
 	http.ResponseWriter
+
 	statusCode int
 	written    bool
 }
@@ -323,6 +237,7 @@ func (sw *statusWriter) WriteHeader(statusCode int) {
 		sw.statusCode = statusCode
 		sw.written = true
 	}
+
 	sw.ResponseWriter.WriteHeader(statusCode)
 }
 
@@ -362,11 +277,12 @@ func RequestLoggerMiddleware(logger *slog.Logger) func(http.Handler) http.Handle
 			// Slow query threshold (1 second)
 			const slowQueryThreshold = 1 * time.Second
 
-			if sw.statusCode >= http.StatusInternalServerError {
+			switch {
+			case sw.statusCode >= http.StatusInternalServerError:
 				logLevel = slog.LevelError
-			} else if sw.statusCode >= http.StatusBadRequest {
+			case sw.statusCode >= http.StatusBadRequest:
 				logLevel = slog.LevelWarn
-			} else if duration >= slowQueryThreshold {
+			case duration >= slowQueryThreshold:
 				logLevel = slog.LevelWarn
 			}
 
@@ -381,3 +297,118 @@ func RequestLoggerMiddleware(logger *slog.Logger) func(http.Handler) http.Handle
 		})
 	}
 }
+
+// DecodedPathMiddleware makes the router match routes against the decoded
+// path, so that the handlers read their URL parameters decoded. Without it,
+// chi matches against the escaped path whenever the request has one and hands
+// the handlers their parameters still escaped: /ip/2001%3Adb8%3A%3A1 would
+// look up "2001%3Adb8%3A%3A1".
+func DecodedPathMiddleware(next http.Handler) http.Handler {
+	return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		chi.RouteContext(r.Context()).RoutePath = r.URL.Path
+
+		next.ServeHTTP(w, r)
+	})
+}
+
+// The values of the headers SecurityHeadersMiddleware sets.
+const (
+	// strictTransportSecurity tells browsers to reach this host and its
+	// subdomains only over HTTPS for a year. The server listens on plain
+	// HTTP behind a TLS-terminating proxy, which passes the header on.
+	strictTransportSecurity = "max-age=31536000; includeSubDomains"
+
+	// contentSecurityPolicy lets a page load content only from this server,
+	// and run or apply an inline script or style block only when the block
+	// carries the nonce made for that answer, which replaces %[1]s. No other
+	// site may frame the pages.
+	contentSecurityPolicy = "default-src 'self'; " +
+		"script-src 'nonce-%[1]s'; style-src 'nonce-%[1]s'; " +
+		"frame-ancestors 'none'; base-uri 'none'; form-action 'self'"
+
+	frameOptions       = "DENY"
+	contentTypeOptions = "nosniff"
+	referrerPolicy     = "strict-origin-when-cross-origin"
+
+	// permissionsPolicy turns off browser features the pages do not use.
+	permissionsPolicy = "accelerometer=(), camera=(), geolocation=(), " +
+		"gyroscope=(), magnetometer=(), microphone=(), payment=(), usb=()"
+)
+
+// cspNonceKey is the request context key of the nonce SecurityHeadersMiddleware
+// made for the answer.
+type cspNonceKey struct{}
+
+// SecurityHeadersMiddleware sets the security headers REPO_POLICIES.md requires
+// on every answer. Each answer's Content-Security-Policy names a nonce of its
+// own; a page reads it with cspNonce and writes it on its script and style
+// blocks.
+func SecurityHeadersMiddleware(next http.Handler) http.Handler {
+	return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		nonce := rand.Text()
+
+		header := w.Header()
+		header.Set("Strict-Transport-Security", strictTransportSecurity)
+		header.Set("Content-Security-Policy",
+			fmt.Sprintf(contentSecurityPolicy, nonce))
+		header.Set("X-Frame-Options", frameOptions)
+		header.Set("X-Content-Type-Options", contentTypeOptions)
+		header.Set("Referrer-Policy", referrerPolicy)
+		header.Set("Permissions-Policy", permissionsPolicy)
+
+		ctx := context.WithValue(r.Context(), cspNonceKey{}, nonce)
+		next.ServeHTTP(w, r.WithContext(ctx))
+	})
+}
+
+// cspNonce returns the nonce SecurityHeadersMiddleware made for the answer to
+// the request with ctx.
+func cspNonce(ctx context.Context) string {
+	nonce, _ := ctx.Value(cspNonceKey{}).(string)
+
+	return nonce
+}
+
+// RealIPMiddleware gives a request from one of trustedProxies the address of
+// the client the proxies forwarded it for as its RemoteAddr: the last
+// X-Forwarded-For entry that is not a trusted proxy. Each proxy appends the
+// address it took the request from, so the entries left of that one are
+// whatever the client sent. A request from any other peer keeps the peer's
+// address, whatever headers it carries.
+func RealIPMiddleware(trustedProxies []netip.Prefix) func(http.Handler) http.Handler {
+	isTrusted := func(addr netip.Addr) bool {
+		return slices.ContainsFunc(trustedProxies, func(proxy netip.Prefix) bool {
+			return proxy.Contains(addr.Unmap())
+		})
+	}
+
+	return func(next http.Handler) http.Handler {
+		return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+			peer, err := netip.ParseAddrPort(r.RemoteAddr)
+			if err != nil || !isTrusted(peer.Addr()) {
+				next.ServeHTTP(w, r)
+
+				return
+			}
+
+			// A header sent on several lines is one list.
+			entries := strings.Split(
+				strings.Join(r.Header.Values("X-Forwarded-For"), ","), ",")
+
+			for _, entry := range slices.Backward(entries) {
+				addr, err := netip.ParseAddr(strings.TrimSpace(entry))
+				if err != nil {
+					break
+				}
+
+				r.RemoteAddr = addr.String()
+
+				if !isTrusted(addr) {
+					break
+				}
+			}
+
+			next.ServeHTTP(w, r)
+		})
+	}
+}

internal/server/middleware_internal_test.go

@@ -0,0 +1,234 @@
+package server
+
+import (
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"net/netip"
+	"strings"
+	"testing"
+	"time"
+)
+
+// The addresses the forwarding tests use: the client a proxy names in
+// X-Forwarded-For, the range of trusted proxies, a peer inside it, and a peer
+// outside it, which lies in testPrefixV4 so that an IP lookup of it finds a
+// route.
+const (
+	forwardedClient   = "203.0.113.9"
+	trustedProxyRange = "10.0.0.0/8"
+	trustedPeer       = "10.0.0.2:1234"
+	untrustedPeerAddr = "192.0.2.1"
+	untrustedPeer     = untrustedPeerAddr + ":1234"
+)
+
+// TestResponsesCarrySecurityHeaders checks that a page, a JSON route and a
+// path no route matches each answer with the security headers REPO_POLICIES.md
+// requires, and that the Content-Security-Policy allows no inline content
+// without a nonce (https://git.eeqj.de/sneak/routewatch/issues/108).
+func TestResponsesCarrySecurityHeaders(t *testing.T) {
+	t.Parallel()
+
+	s := newServerWithTestRoutes(t)
+
+	headers := []string{
+		"Strict-Transport-Security",
+		"Content-Security-Policy",
+		"X-Frame-Options",
+		"X-Content-Type-Options",
+		"Referrer-Policy",
+		"Permissions-Policy",
+	}
+
+	for _, path := range []string{"/", statsPath, "/no-such-page"} {
+		rec := httptest.NewRecorder()
+		s.router.ServeHTTP(rec, httptest.NewRequestWithContext(
+			t.Context(), http.MethodGet, path, nil))
+
+		for _, name := range headers {
+			if rec.Header().Get(name) == "" {
+				t.Errorf("%s: no %s header", path, name)
+			}
+		}
+
+		policy := rec.Header().Get("Content-Security-Policy")
+		if strings.Contains(policy, "'unsafe-") {
+			t.Errorf("%s: Content-Security-Policy %q allows unsafe content",
+				path, policy)
+		}
+	}
+}
+
+// TestJSONValidationKeeps4xxWithoutBody wraps a handler that answers 400
+// without a body in JSONValidationMiddleware and checks that the JSON error
+// body keeps the 400 and gives the status's own text. Before the fix it said
+// "Internal server error"
+// (https://git.eeqj.de/sneak/routewatch/issues/111).
+func TestJSONValidationKeeps4xxWithoutBody(t *testing.T) {
+	t.Parallel()
+
+	handler := JSONValidationMiddleware(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			w.WriteHeader(http.StatusBadRequest)
+		}))
+
+	rec := httptest.NewRecorder()
+	handler.ServeHTTP(rec, httptest.NewRequestWithContext(
+		t.Context(), http.MethodGet, "/", nil))
+
+	const name = "400 without a body"
+
+	if rec.Code != http.StatusBadRequest {
+		t.Errorf("%s: status %d, want %d; body %s",
+			name, rec.Code, http.StatusBadRequest, rec.Body)
+	}
+
+	checkJSONErrorCode(t, name, rec, http.StatusBadRequest)
+	checkJSONErrorMessage(t, name, rec, http.StatusText(http.StatusBadRequest))
+}
+
+// TestPagesCarryTheCSPNonce checks that every script and style block of each
+// page carries the nonce its Content-Security-Policy header names, so that the
+// browser runs and applies them, that each answer has a nonce of its own, and
+// that no page uses a style attribute, which the policy does not allow.
+func TestPagesCarryTheCSPNonce(t *testing.T) {
+	t.Parallel()
+
+	s := newServerWithTestRoutes(t)
+
+	_, err := s.db.GetOrCreateASN(64500, time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC))
+	if err != nil {
+		t.Fatalf("GetOrCreateASN: %v", err)
+	}
+
+	paths := []string{
+		"/", "/status", asPagePath, "/prefix/" + testPrefixV4,
+		prefixLengthPagePath, "/prefixlength6/32",
+	}
+	seen := make(map[string]bool)
+
+	for _, path := range paths {
+		rec := httptest.NewRecorder()
+		s.router.ServeHTTP(rec, httptest.NewRequestWithContext(
+			t.Context(), http.MethodGet, path, nil))
+
+		if rec.Code != http.StatusOK {
+			t.Errorf("%s: status %d, want %d; body %s",
+				path, rec.Code, http.StatusOK, rec.Body)
+
+			continue
+		}
+
+		policy := rec.Header().Get("Content-Security-Policy")
+
+		_, nonce, found := strings.Cut(policy, "'nonce-")
+		if !found {
+			t.Errorf("%s: Content-Security-Policy %q names no nonce", path, policy)
+
+			continue
+		}
+
+		nonce, _, _ = strings.Cut(nonce, "'")
+		if seen[nonce] {
+			t.Errorf("%s: nonce %q was already used for another answer", path, nonce)
+		}
+
+		seen[nonce] = true
+		page := rec.Body.String()
+
+		for _, tag := range []string{"<script", "<style"} {
+			withNonce := tag + ` nonce="` + nonce + `"`
+			if strings.Count(page, tag) != strings.Count(page, withNonce) {
+				t.Errorf("%s: a %s tag does not carry the nonce %q", path, tag, nonce)
+			}
+		}
+
+		if strings.Contains(page, "style=") {
+			t.Errorf("%s: the page uses a style attribute", path)
+		}
+	}
+}
+
+// TestForwardedForFromUntrustedPeerIsIgnored asks /ip/ for the caller's own
+// address from a peer that is not a trusted proxy, with forwarding headers
+// naming another address, and checks that the lookup is for the TCP peer.
+// Before the fix any client could choose the address it was looked up and
+// logged as (https://git.eeqj.de/sneak/routewatch/issues/108).
+func TestForwardedForFromUntrustedPeerIsIgnored(t *testing.T) {
+	t.Parallel()
+
+	s := newServerWithTestRoutes(t)
+
+	req := httptest.NewRequestWithContext(t.Context(), http.MethodGet, "/ip/", nil)
+	req.RemoteAddr = untrustedPeer
+	req.Header.Set("X-Forwarded-For", forwardedClient)
+	req.Header.Set("X-Real-IP", forwardedClient)
+	req.Header.Set("True-Client-IP", forwardedClient)
+
+	rec := httptest.NewRecorder()
+	s.router.ServeHTTP(rec, req)
+
+	var body struct {
+		Data IPLookupResponse `json:"data"`
+	}
+
+	err := json.Unmarshal(rec.Body.Bytes(), &body)
+	if err != nil {
+		t.Fatalf("decoding the answer %q: %v", rec.Body, err)
+	}
+
+	if body.Data.Query != untrustedPeerAddr {
+		t.Errorf("looked up %q, want the TCP peer %s; answer %s",
+			body.Data.Query, untrustedPeerAddr, rec.Body)
+	}
+}
+
+// TestRealIPMiddlewareTrustsOnlyConfiguredProxies checks which address
+// RealIPMiddleware gives a request: the TCP peer, unless the peer is a trusted
+// proxy; then the last X-Forwarded-For entry that is not a trusted proxy,
+// since the entries left of it are whatever the client sent.
+func TestRealIPMiddlewareTrustsOnlyConfiguredProxies(t *testing.T) {
+	t.Parallel()
+
+	trusted := []netip.Prefix{netip.MustParsePrefix(trustedProxyRange)}
+
+	tests := []struct {
+		name         string
+		peer         string
+		forwardedFor string
+		want         string
+	}{
+		{"untrusted peer", untrustedPeer, forwardedClient, untrustedPeer},
+		{"trusted peer", trustedPeer, forwardedClient, forwardedClient},
+		{
+			"entries the client wrote", trustedPeer,
+			"198.51.100.7, " + forwardedClient, forwardedClient,
+		},
+		{
+			"two trusted proxies", trustedPeer,
+			forwardedClient + ", 10.0.0.3", forwardedClient,
+		},
+		{"no header", trustedPeer, "", trustedPeer},
+		{"entry not an address", trustedPeer, "client", trustedPeer},
+	}
+
+	for _, tt := range tests {
+		var got string
+
+		handler := RealIPMiddleware(trusted)(http.HandlerFunc(
+			func(_ http.ResponseWriter, r *http.Request) { got = r.RemoteAddr }))
+
+		req := httptest.NewRequestWithContext(t.Context(), http.MethodGet, "/", nil)
+		req.RemoteAddr = tt.peer
+
+		if tt.forwardedFor != "" {
+			req.Header.Set("X-Forwarded-For", tt.forwardedFor)
+		}
+
+		handler.ServeHTTP(httptest.NewRecorder(), req)
+
+		if got != tt.want {
+			t.Errorf("%s: RemoteAddr %q, want %q", tt.name, got, tt.want)
+		}
+	}
+}

internal/server/routes.go

@@ -1,30 +1,53 @@
 package server
 
 import (
+	"net/http"
 	"time"
 
 	"github.com/go-chi/chi/v5"
 	"github.com/go-chi/chi/v5/middleware"
 )
 
+// The error messages the JSON routes under /api/v1 and /ip answer with when no
+// route matches the path, and when none takes the method.
+const (
+	notFoundMessage         = "Not found"
+	methodNotAllowedMessage = "Method not allowed"
+
+	// jsonRoutesAllow is the Allow header of their 405 answer: every one of
+	// them takes GET and HEAD, and no other method.
+	jsonRoutesAllow = "GET, HEAD"
+)
+
 // setupRoutes configures the HTTP routes
 func (s *Server) setupRoutes() {
 	r := chi.NewRouter()
 
 	// Middleware
+	r.Use(DecodedPathMiddleware)
+	// GetHead serves HEAD with the GET handler. It looks the route up by the
+	// path DecodedPathMiddleware sets, so it comes after it.
+	r.Use(middleware.GetHead)
+	r.Use(SecurityHeadersMiddleware)
 	r.Use(middleware.RequestID)
-	r.Use(middleware.RealIP)
+	r.Use(RealIPMiddleware(s.trustedProxies))
 	r.Use(RequestLoggerMiddleware(s.logger.Logger)) // Structured request logging
 	r.Use(middleware.Recoverer)
-	const requestTimeout = 30 * time.Second // Increased from 8s for slow queries
-	r.Use(TimeoutMiddleware(requestTimeout, s.logger.Logger))
 	r.Use(JSONResponseMiddleware)
+	// Timeout must come after JSONResponseMiddleware, which sends its answer
+	// when the handler returns and would send a 200 ahead of Timeout's 504 for
+	// a handler that wrote nothing.
+	const requestTimeout = 30 * time.Second // Increased from 8s for slow queries
+	r.Use(middleware.Timeout(requestTimeout))
 
 	// Routes
 	r.Get("/", s.handleIndex())
 	r.Get("/status", s.handleStatusHTML())
 	r.Get("/status.json", JSONValidationMiddleware(s.handleStatusJSON()).ServeHTTP)
-	r.Get("/.well-known/healthcheck.json", JSONValidationMiddleware(s.handleHealthCheck()).ServeHTTP)
+	r.Get(
+		"/.well-known/healthcheck.json",
+		JSONValidationMiddleware(s.handleHealthCheck()).ServeHTTP,
+	)
 
 	// AS and prefix detail pages
 	r.Get("/as/{asn}", s.handleASDetail())
@@ -32,20 +55,39 @@ func (s *Server) setupRoutes() {
 	r.Get("/prefixlength/{length}", s.handlePrefixLength())
 	r.Get("/prefixlength6/{length}", s.handlePrefixLength6())
 
+	// The JSON routes answer a path or method they do not take with the JSON
+	// error body, as they answer everything else.
+	jsonNotFound := func(w http.ResponseWriter, _ *http.Request) {
+		s.writeJSONError(w, http.StatusNotFound, notFoundMessage)
+	}
+	jsonMethodNotAllowed := func(w http.ResponseWriter, _ *http.Request) {
+		// chi's own 405 answer sets Allow; one set in its place must too.
+		w.Header().Set("Allow", jsonRoutesAllow)
+		s.writeJSONError(w, http.StatusMethodNotAllowed, methodNotAllowedMessage)
+	}
+
 	// IP info JSON endpoints (replaces old /ip redirect)
 	r.Route("/ip", func(r chi.Router) {
 		r.Use(JSONValidationMiddleware)
+		r.NotFound(jsonNotFound)
+		r.MethodNotAllowed(jsonMethodNotAllowed)
 		r.Get("/", s.handleIPInfo()) // Client IP
+		// GetHead does not reach this route: mounting the group registers
+		// /ip and /ip/ for every method, so HEAD arrives here as HEAD.
+		r.Head("/", s.handleIPInfo())
 		r.Get("/{addr}", s.handleIPInfo()) // Specified IP
 	})
 
 	// API routes
 	r.Route("/api/v1", func(r chi.Router) {
 		r.Use(JSONValidationMiddleware)
+		r.NotFound(jsonNotFound)
+		r.MethodNotAllowed(jsonMethodNotAllowed)
 		r.Get("/stats", s.handleStats())
 		r.Get("/ip/{ip}", s.handleIPLookup())
 		r.Get("/as/{asn}", s.handleASDetailJSON())
 		r.Get("/prefix/{prefix}/{len}", s.handlePrefixDetailJSON())
+		r.Get("/search/as", s.handleASSearch())
 	})
 
 	s.router = r

internal/server/routes_internal_test.go

@@ -0,0 +1,100 @@
+package server
+
+import (
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+)
+
+// TestRoutesAnswerHEAD sends HEAD for a page and JSON routes through the
+// server's router and checks that each answers as GET does. /ip and /ip/, the
+// lookup of the caller's own address, are in the list because chi's GetHead
+// does not reach them. Before the fix every route answered HEAD with 405
+// (https://git.eeqj.de/sneak/routewatch/issues/111).
+func TestRoutesAnswerHEAD(t *testing.T) {
+	t.Parallel()
+
+	s := newServerWithTestRoutes(t)
+
+	for _, path := range []string{"/", statsPath, "/ip", "/ip/"} {
+		rec := httptest.NewRecorder()
+		s.router.ServeHTTP(rec, httptest.NewRequestWithContext(
+			t.Context(), http.MethodHead, path, nil))
+
+		if rec.Code != http.StatusOK {
+			t.Errorf("HEAD %s: status %d, want %d; body %s",
+				path, rec.Code, http.StatusOK, rec.Body)
+		}
+	}
+}
+
+// TestJSONRoutesAnswerUnmatchedRequestsWithJSONErrors sends a method the JSON
+// routes do not take, and a path none of them matches, to each group of JSON
+// routes through the server's router. Each must answer with the JSON error
+// body, and a 405 must name the methods the routes take in its Allow header.
+// Before the fix a 405 said "Internal server error" and a 404 was chi's plain
+// text answer (https://git.eeqj.de/sneak/routewatch/issues/111).
+func TestJSONRoutesAnswerUnmatchedRequestsWithJSONErrors(t *testing.T) {
+	t.Parallel()
+
+	s := newServerWithTestRoutes(t)
+
+	tests := []struct {
+		method   string
+		path     string
+		wantCode int
+		wantMsg  string
+	}{
+		{http.MethodPost, statsPath, http.StatusMethodNotAllowed, methodNotAllowedMessage},
+		{http.MethodPost, "/ip/", http.StatusMethodNotAllowed, methodNotAllowedMessage},
+		{http.MethodGet, "/api/v1/nothing", http.StatusNotFound, notFoundMessage},
+		{http.MethodGet, "/ip/192.0.2.1/nothing", http.StatusNotFound, notFoundMessage},
+	}
+
+	for _, tt := range tests {
+		name := tt.method + " " + tt.path
+
+		rec := httptest.NewRecorder()
+		s.router.ServeHTTP(rec, httptest.NewRequestWithContext(
+			t.Context(), tt.method, tt.path, nil))
+
+		if rec.Code != tt.wantCode {
+			t.Errorf("%s: status %d, want %d; body %s",
+				name, rec.Code, tt.wantCode, rec.Body)
+		}
+
+		checkJSONErrorCode(t, name, rec, tt.wantCode)
+		checkJSONErrorMessage(t, name, rec, tt.wantMsg)
+
+		allow := rec.Header().Get("Allow")
+		if tt.wantCode == http.StatusMethodNotAllowed && allow != jsonRoutesAllow {
+			t.Errorf("%s: Allow %q, want %q", name, allow, jsonRoutesAllow)
+		}
+	}
+}
+
+// checkJSONErrorMessage checks that rec holds the JSON error body with the
+// given message.
+func checkJSONErrorMessage(
+	t *testing.T, name string, rec *httptest.ResponseRecorder, want string,
+) {
+	t.Helper()
+
+	var body struct {
+		Error struct {
+			Msg string `json:"msg"`
+		} `json:"error"`
+	}
+
+	err := json.Unmarshal(rec.Body.Bytes(), &body)
+	if err != nil {
+		t.Errorf("%s: decoding the answer %q: %v", name, rec.Body, err)
+
+		return
+	}
+
+	if body.Error.Msg != want {
+		t.Errorf("%s: error message %q, want %q", name, body.Error.Msg, want)
+	}
+}

internal/server/server.go

@@ -1,12 +1,21 @@
-// Package server provides HTTP endpoints for status monitoring and statistics
+// Package server serves RouteWatch's web pages and JSON API:
+//   - the home page, the status figures and the health check
+//   - the AS, prefix and prefix length pages, and the JSON answers behind them
+//   - the IP lookups
+//   - the AS name search
 package server
 
 import (
 	"context"
+	"errors"
+	"fmt"
+	"net"
 	"net/http"
+	"net/netip"
 	"strconv"
 	"time"
 
+	"git.eeqj.de/sneak/routewatch/internal/blocklist"
 	"git.eeqj.de/sneak/routewatch/internal/config"
 	"git.eeqj.de/sneak/routewatch/internal/database"
 	"git.eeqj.de/sneak/routewatch/internal/logger"
@@ -28,6 +37,20 @@ type ASNFetcher interface {
 	GetStats() ASNFetcherStats
 }
 
+// Blocklists tells which public blocklists list an address, from copies it
+// keeps in memory, as *blocklist.Blocklists does.
+type Blocklists interface {
+	Lookup(addr netip.Addr) []blocklist.Listing
+	ShortestInterval() time.Duration
+}
+
+// dnsResolver looks up the addresses of host names and the names of addresses
+// in DNS, as *net.Resolver does.
+type dnsResolver interface {
+	LookupHost(ctx context.Context, host string) ([]string, error)
+	LookupAddr(ctx context.Context, addr string) ([]string, error)
+}
+
 // Server provides HTTP endpoints for status monitoring
 type Server struct {
 	router     *chi.Mux
@@ -37,15 +60,31 @@ type Server struct {
 	port       int
 	srv        *http.Server
 	asnFetcher ASNFetcher
+	blocklists Blocklists
+
+	// trustedProxies are the reverse proxies whose X-Forwarded-For header
+	// RealIPMiddleware believes
+	trustedProxies []netip.Prefix
+
+	// resolver makes the IP lookup's DNS lookups: net.DefaultResolver, or a
+	// fake in tests
+	resolver dnsResolver
 }
 
 // New creates a new HTTP server
-func New(db database.Store, streamer *streamer.Streamer, logger *logger.Logger, cfg *config.Config) *Server {
+func New(
+	db database.Store,
+	streamer *streamer.Streamer,
+	logger *logger.Logger,
+	cfg *config.Config,
+) *Server {
 	s := &Server{
 		db:             db,
 		streamer:       streamer,
 		logger:         logger,
 		port:           cfg.Port,
+		trustedProxies: cfg.TrustedProxies,
+		resolver:       net.DefaultResolver,
 	}
 
 	s.setupRoutes()
@@ -53,9 +92,12 @@ func New(db database.Store, streamer *streamer.Streamer, logger *logger.Logger,
 	return s
 }
 
-// Start starts the HTTP server
+// Start listens on the configured port and serves HTTP in the background.
-func (s *Server) Start() error {
+// It returns an error if it cannot listen, for example because the port is
+// already in use. ctx bounds only opening the port, not serving.
+func (s *Server) Start(ctx context.Context) error {
 	const (
+		network           = "tcp"
 		readHeaderTimeout = 40 * time.Second
 		readTimeout       = 60 * time.Second
 		writeTimeout      = 60 * time.Second
@@ -73,10 +115,18 @@ func (s *Server) Start() error {
 
 	s.logger.Info("Starting HTTP server", "port", s.port, "addr", s.srv.Addr)
 
-	// Start in goroutine but log when actually listening
+	var listenConfig net.ListenConfig
+
+	listener, err := listenConfig.Listen(ctx, network, s.srv.Addr)
+	if err != nil {
+		return fmt.Errorf("starting HTTP server: %w", err)
+	}
+
+	s.logger.Info("HTTP server listening", "addr", listener.Addr().String())
+
 	go func() {
-		s.logger.Info("HTTP server listening", "addr", s.srv.Addr)
+		err := s.srv.Serve(listener)
-		if err := s.srv.ListenAndServe(); err != nil && err != http.ErrServerClosed {
+		if err != nil && !errors.Is(err, http.ErrServerClosed) {
 			s.logger.Error("HTTP server error", "error", err)
 		}
 	}()
@@ -84,7 +134,9 @@ func (s *Server) Start() error {
 	return nil
 }
 
-// Stop gracefully stops the HTTP server
+// Stop stops the HTTP server. It waits, until ctx ends, for the requests in
+// progress to finish, and then closes the connections of those still running
+// and returns ctx's error.
 func (s *Server) Stop(ctx context.Context) error {
 	if s.srv == nil {
 		return nil
@@ -92,10 +144,22 @@ func (s *Server) Stop(ctx context.Context) error {
 
 	s.logger.Info("Stopping HTTP server")
 
-	return s.srv.Shutdown(ctx)
+	err := s.srv.Shutdown(ctx)
+	if err != nil {
+		return errors.Join(err, s.srv.Close())
 	}
 
-// SetASNFetcher sets the ASN WHOIS fetcher for on-demand lookups.
+	return nil
+}
+
+// SetASNFetcher sets the ASN WHOIS fetcher for on-demand lookups. Call it
+// before Start: the handlers read the fetcher without a lock.
 func (s *Server) SetASNFetcher(fetcher ASNFetcher) {
 	s.asnFetcher = fetcher
 }
+
+// SetBlocklists sets the blocklists the IP lookups report. Call it before
+// Start: the handlers read the blocklists without a lock.
+func (s *Server) SetBlocklists(blocklists Blocklists) {
+	s.blocklists = blocklists
+}

internal/server/server_internal_test.go

@@ -0,0 +1,92 @@
+package server
+
+import (
+	"context"
+	"net"
+	"net/http"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+)
+
+// TestStopClosesTheConnectionOfARequestStillRunning serves a request whose
+// handler does not return until the test ends, and stops the server with a
+// deadline that passes while the request runs. The client must then see its
+// connection closed. Before the fix Stop returned at the deadline and left the
+// connection open until the handler returned.
+func TestStopClosesTheConnectionOfARequestStillRunning(t *testing.T) {
+	t.Parallel()
+
+	const (
+		stopDeadline = 100 * time.Millisecond
+		// waitTimeout is how long the test waits for the request to reach
+		// the handler, and then for the client to see its connection closed.
+		waitTimeout = 2 * time.Second
+		// readHeaderTimeout bounds the test server's read of the request
+		// header, as every http.Server must.
+		readHeaderTimeout = time.Second
+	)
+
+	release := make(chan struct{})
+	defer close(release)
+
+	started := make(chan struct{})
+
+	var listenConfig net.ListenConfig
+
+	listener, err := listenConfig.Listen(t.Context(), "tcp", "127.0.0.1:0")
+	if err != nil {
+		t.Fatalf("failed to open the test listener: %v", err)
+	}
+
+	s := &Server{logger: logger.New(), srv: &http.Server{
+		Handler: http.HandlerFunc(func(http.ResponseWriter, *http.Request) {
+			close(started)
+			<-release
+		}),
+		ReadHeaderTimeout: readHeaderTimeout,
+	}}
+
+	go func() { _ = s.srv.Serve(listener) }()
+
+	answered := make(chan error, 1)
+
+	go func() {
+		req, err := http.NewRequestWithContext(t.Context(), http.MethodGet,
+			"http://"+listener.Addr().String()+"/", nil)
+		if err != nil {
+			answered <- err
+
+			return
+		}
+
+		resp, err := http.DefaultClient.Do(req)
+		if err == nil {
+			_ = resp.Body.Close()
+		}
+
+		answered <- err
+	}()
+
+	select {
+	case <-started:
+	case <-time.After(waitTimeout):
+		t.Fatal("the request did not reach the handler")
+	}
+
+	ctx, cancel := context.WithTimeout(t.Context(), stopDeadline)
+	defer cancel()
+
+	_ = s.Stop(ctx)
+
+	select {
+	case err := <-answered:
+		if err == nil {
+			t.Error("the request still running at the deadline was answered")
+		}
+	case <-time.After(waitTimeout):
+		t.Fatal("the connection of the request still running at the " +
+			"deadline was open after Stop returned")
+	}
+}

internal/server/server_test.go

@@ -0,0 +1,47 @@
+package server_test
+
+import (
+	"context"
+	"net"
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/config"
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/metrics"
+	"git.eeqj.de/sneak/routewatch/internal/server"
+	"git.eeqj.de/sneak/routewatch/internal/streamer"
+)
+
+// TestStartFailsWhenPortIsHeld holds a port with a listener of its own and
+// checks that Start on that port returns an error. Before the fix Start
+// listened in a goroutine, only logged the failure and returned nil, so the
+// daemon kept running without its web server.
+func TestStartFailsWhenPortIsHeld(t *testing.T) {
+	t.Parallel()
+
+	var listenConfig net.ListenConfig
+
+	held, err := listenConfig.Listen(t.Context(), "tcp", ":0")
+	if err != nil {
+		t.Fatalf("failed to open the test listener: %v", err)
+	}
+	defer func() { _ = held.Close() }()
+
+	addr, ok := held.Addr().(*net.TCPAddr)
+	if !ok {
+		t.Fatalf("test listener address %v is not a TCP address", held.Addr())
+	}
+
+	port := addr.Port
+	log := logger.New()
+	s := server.New(nil, streamer.New(log, metrics.New()), log,
+		&config.Config{Port: port})
+
+	defer func() { _ = s.Stop(context.Background()) }()
+
+	err = s.Start(t.Context())
+	if err == nil {
+		t.Fatalf("Start returned nil while port %d was held by another listener",
+			port)
+	}
+}

internal/streamer/streamer.go

@@ -1,5 +1,6 @@
-// Package streamer implements an HTTP client that connects to the RIPE RIS Live streaming API,
+// Package streamer implements an HTTP client that connects to the RIPE RIS Live
-// parses BGP UPDATE messages from the JSON stream, and dispatches them to registered handlers.
+// streaming API, parses BGP UPDATE messages from the JSON stream, and
+// dispatches them to registered handlers.
 package streamer
 
 import (
@@ -7,6 +8,7 @@ import (
 	"compress/gzip"
 	"context"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"io"
 	"math"
@@ -24,20 +26,20 @@ import (
 // countingReader wraps an io.Reader and counts bytes read
 type countingReader struct {
 	reader io.Reader
-	count  int64
+	count  atomic.Int64
 }
 
 // Read implements io.Reader and counts bytes
 func (c *countingReader) Read(p []byte) (int, error) {
 	n, err := c.reader.Read(p)
-	atomic.AddInt64(&c.count, int64(n))
+	c.count.Add(int64(n))
 
 	return n, err
 }
 
 // Count returns the total bytes read
 func (c *countingReader) Count() int64 {
-	return atomic.LoadInt64(&c.count)
+	return c.count.Load()
 }
 
 // Configuration constants for the RIS Live streamer.
@@ -48,9 +50,13 @@ const (
 	metricsUpdateRate = time.Second
 	minBackoffDelay   = 5 * time.Second
 	maxBackoffDelay   = 320 * time.Second
+	// Reconnect when no line arrives for this long
+	streamIdleTimeout  = 60 * time.Second
 	metricsLogInterval = 10 * time.Second
 	bytesPerKB         = 1024
 	bytesPerMB         = 1024 * 1024
+	// Well above an extended BGP message (RFC 8654) in hex plus its prefixes
+	maxLineSize           = bytesPerMB
 	maxConcurrentHandlers = 800 // Maximum number of concurrent message handlers
 
 	// Backpressure constants
@@ -58,11 +64,26 @@ const (
 	backpressureSlope     = 2.0 // Slope for linear drop probability increase
 )
 
+// messageTypeNotification is the type of a BGP NOTIFICATION message, which ends
+// the session with the peer.
+const messageTypeNotification = "NOTIFICATION"
+
+var (
+	errAlreadyRunning   = errors.New("streamer already running")
+	errUnexpectedStatus = errors.New("unexpected status code")
+	// errStreamEnded is returned by stream when RIS Live ends the response.
+	errStreamEnded = errors.New("RIS Live ended the stream")
+	// errStreamIdle is returned by stream when no line arrived for the idle
+	// timeout.
+	errStreamIdle = errors.New("no line received")
+)
+
 // MessageHandler defines the interface for processing RIS messages.
 // Implementations must specify which message types they want to receive,
 // how to process messages, and their desired queue capacity.
 type MessageHandler interface {
-	// WantsMessage returns true if this handler wants to process messages of the given type.
+	// WantsMessage returns true if this handler wants to process messages of
+	// the given type.
 	WantsMessage(messageType string) bool
 
 	// HandleMessage processes a RIS message. This method is called from a dedicated
@@ -73,7 +94,8 @@ type MessageHandler interface {
 	// QueueCapacity returns the desired queue capacity for this handler.
 	// Handlers that process quickly can have larger queues to buffer bursts.
 	// When the queue fills up, messages will be dropped according to the
-	// backpressure algorithm.
+	// backpressure algorithm, except STATE messages, which are held until
+	// there is room.
 	QueueCapacity() int
 }
 
@@ -84,12 +106,12 @@ type RawMessageHandler func(line string)
 // handlerMetrics tracks performance metrics for a handler
 type handlerMetrics struct {
 	processedCount     uint64        // Total messages processed
-	droppedCount       uint64        // Total messages dropped
+	droppedCount       atomic.Uint64 // Total messages dropped; not under mu
 	totalTime          time.Duration // Total processing time (for average calculation)
 	minTime            time.Duration // Minimum processing time
 	maxTime            time.Duration // Maximum processing time
 	queueHighWaterMark int           // Maximum queue length seen
-	mu                 sync.Mutex    // Protects the metrics
+	mu                 sync.Mutex    // Protects the other metrics
 }
 
 // handlerInfo wraps a handler with its queue and metrics
@@ -97,6 +119,9 @@ type handlerInfo struct {
 	handler MessageHandler
 	queue   chan *ristypes.RISMessage
 	metrics handlerMetrics
+	// heldState holds, in feed order, the STATE messages that found queue
+	// full (see queueMessage). Only the read loop uses it, and then Stop.
+	heldState []*ristypes.RISMessage
 }
 
 // Streamer manages a connection to the RIPE RIS Live streaming API for receiving
@@ -107,13 +132,18 @@ type Streamer struct {
 	logger *logger.Logger
 	client *http.Client
 	url    string
+	// Drop the connection when no line arrives for this long
+	idleTimeout time.Duration
+	// First wait before reconnecting; doubles up to maxBackoffDelay
+	minBackoff   time.Duration
 	handlers     []*handlerInfo
 	rawHandler   RawMessageHandler
 	mu           sync.RWMutex
 	cancel       context.CancelFunc
 	running      bool
+	workers      sync.WaitGroup // the handler workers; Stop waits for them
 	metrics      *metrics.Tracker
-	totalDropped uint64          // Total dropped messages across all handlers
+	totalDropped atomic.Uint64   // Total dropped messages across all handlers
 	random       *rand.Rand      // Random number generator for backpressure drops
 	bgpPeers     map[string]bool // Track active BGP peers by peer IP
 	bgpPeersMu   sync.RWMutex    // Protects bgpPeers map
@@ -121,11 +151,14 @@ type Streamer struct {
 
 // New creates a new Streamer instance configured to connect to the RIS Live API.
 // The logger is used for structured logging of connection events and errors.
-// The metrics tracker is used to record message counts, bytes received, and connection status.
+// The metrics tracker is used to record message counts, bytes received, and
+// connection status.
 func New(logger *logger.Logger, metrics *metrics.Tracker) *Streamer {
 	return &Streamer{
 		logger:      logger,
 		url:         risLiveURL,
+		idleTimeout: streamIdleTimeout,
+		minBackoff:  minBackoffDelay,
 		client: &http.Client{
 			Timeout: 0, // No timeout for streaming
 			Transport: &http.Transport{
@@ -153,7 +186,8 @@ func (s *Streamer) RegisterHandler(handler MessageHandler) {
 		handler: handler,
 		queue:   make(chan *ristypes.RISMessage, handler.QueueCapacity()),
 		metrics: handlerMetrics{
-			minTime: time.Duration(math.MaxInt64), // Initialize to max so first value sets the floor
+			// Initialize to max so first value sets the floor
+			minTime: time.Duration(math.MaxInt64),
 		},
 	}
 
@@ -161,37 +195,49 @@ func (s *Streamer) RegisterHandler(handler MessageHandler) {
 
 	// If we're already running, start a worker for this handler
 	if s.running {
+		s.workers.Add(1)
+
 		go s.runHandlerWorker(info)
 	}
 }
 
+// SetURL makes the streamer read from url instead of RIS Live. Call it before
+// Start. Tests use it to read from a local server.
+func (s *Streamer) SetURL(url string) {
+	s.url = url
+}
+
 // RegisterRawHandler sets a callback to receive raw JSON lines from the stream
 // before they are parsed. Only one raw handler can be registered at a time;
 // subsequent calls will replace the previous handler.
 func (s *Streamer) RegisterRawHandler(handler RawMessageHandler) {
 	s.mu.Lock()
 	defer s.mu.Unlock()
+
 	s.rawHandler = handler
 }
 
 // Start begins streaming BGP updates from the RIS Live API in a background goroutine.
 // It starts worker goroutines for each registered handler and manages automatic
 // reconnection with exponential backoff on connection failures.
-// Returns an error if the streamer is already running.
+// Returns an error if the streamer is already running. It streams until Stop:
-func (s *Streamer) Start() error {
+// the end of ctx does not stop it.
+func (s *Streamer) Start(ctx context.Context) error {
 	s.mu.Lock()
 	defer s.mu.Unlock()
 
 	if s.running {
-		return fmt.Errorf("streamer already running")
+		return errAlreadyRunning
 	}
 
-	ctx, cancel := context.WithCancel(context.Background())
+	ctx, cancel := context.WithCancel(context.WithoutCancel(ctx))
 	s.cancel = cancel
 	s.running = true
 
 	// Start workers for each handler
 	for _, info := range s.handlers {
+		s.workers.Add(1)
+
 		go s.runHandlerWorker(info)
 	}
 
@@ -205,10 +251,12 @@ func (s *Streamer) Start() error {
 	return nil
 }
 
-// Stop halts the streaming connection and shuts down all handler workers.
+// Stop ends the stream and stops the handler workers. It returns once each
-// It cancels the streaming context, closes all handler queues, and updates
+// worker has handed its handler every message left on its queue and then the
-// the connection status in metrics. This method is safe to call multiple times.
+// STATE messages held for it (see queueMessage), or once ctx ends, and then
-func (s *Streamer) Stop() {
+// logs how many messages were not handed over. This method is safe to call
+// multiple times.
+func (s *Streamer) Stop(ctx context.Context) {
 	s.mu.Lock()
 	if s.cancel == nil {
 		// Not started, or already stopped: closing the queues again would panic.
@@ -216,44 +264,61 @@ func (s *Streamer) Stop() {
 
 		return
 	}
+
+	// Once the stream's context is cancelled the read loop queues and holds
+	// nothing more (see dispatchMessage).
 	s.cancel()
 	s.cancel = nil
-	// Close all handler queues to signal workers to stop
+
+	notHandedOver := 0
+
 	for _, info := range s.handlers {
+		// The held messages came after everything on the queue. Each waits
+		// for the worker, which does not take s.mu, to make room, until ctx
+		// ends.
+		for len(info.heldState) > 0 && ctx.Err() == nil {
+			select {
+			case info.queue <- info.heldState[0]:
+				info.heldState = info.heldState[1:]
+			case <-ctx.Done():
+			}
+		}
+
+		notHandedOver += len(info.heldState)
+		info.heldState = nil
+
+		// The worker returns once it has emptied the closed queue.
 		close(info.queue)
 	}
+
+	handlers := s.handlers
 	s.running = false
 	s.mu.Unlock()
 	s.metrics.SetConnected(false)
-}
 
-// runHandlerWorker processes messages for a specific handler
+	workersDone := make(chan struct{})
-func (s *Streamer) runHandlerWorker(info *handlerInfo) {
-	for msg := range info.queue {
-		start := time.Now()
-		info.handler.HandleMessage(msg)
-		elapsed := time.Since(start)
 
-		// Update metrics
+	go func() {
-		info.metrics.mu.Lock()
+		s.workers.Wait()
-		info.metrics.processedCount++
+		close(workersDone)
-		info.metrics.totalTime += elapsed
+	}()
 
-		// Update min time
+	select {
-		if elapsed < info.metrics.minTime {
+	case <-workersDone:
-			info.metrics.minTime = elapsed
+	case <-ctx.Done():
-		}
+		for _, info := range handlers {
-
+			notHandedOver += len(info.queue)
-		// Update max time
-		if elapsed > info.metrics.maxTime {
-			info.metrics.maxTime = elapsed
-		}
-		info.metrics.mu.Unlock()
 		}
 	}
 
-// IsRunning reports whether the streamer is currently connected and processing messages.
+	if notHandedOver > 0 {
-// This is safe to call concurrently from multiple goroutines.
+		s.logger.Error("Stop deadline passed before the handlers took every "+
+			"message", "messages", notHandedOver)
+	}
+}
+
+// IsRunning reports whether the streamer is currently connected and processing
+// messages. This is safe to call concurrently from multiple goroutines.
 func (s *Streamer) IsRunning() bool {
 	s.mu.RLock()
 	defer s.mu.RUnlock()
@@ -306,7 +371,7 @@ func (s *Streamer) GetHandlerStats() []HandlerStats {
 			QueueCapacity:      cap(info.queue),
 			QueueHighWaterMark: info.metrics.queueHighWaterMark,
 			ProcessedCount:     info.metrics.processedCount,
-			DroppedCount:       info.metrics.droppedCount,
+			DroppedCount:       info.metrics.droppedCount.Load(),
 			MinProcessTime:     info.metrics.minTime,
 			MaxProcessTime:     info.metrics.maxTime,
 		}
@@ -314,12 +379,15 @@ func (s *Streamer) GetHandlerStats() []HandlerStats {
 		// Calculate average time
 		if info.metrics.processedCount > 0 {
 			processedCount := info.metrics.processedCount
+
 			const maxInt64 = 1<<63 - 1
+
 			if processedCount > maxInt64 {
 				processedCount = maxInt64
 			}
-			//nolint:gosec // processedCount is explicitly bounded above
+
-			hs.AvgProcessTime = info.metrics.totalTime / time.Duration(processedCount)
+			hs.AvgProcessTime = info.metrics.totalTime /
+				time.Duration(processedCount)
 		}
 
 		info.metrics.mu.Unlock()
@@ -330,11 +398,40 @@ func (s *Streamer) GetHandlerStats() []HandlerStats {
 	return stats
 }
 
-// GetDroppedMessages returns the total number of messages dropped across all handlers
+// GetDroppedMessages returns the total number of messages dropped across all
-// due to queue overflow or backpressure. This counter is monotonically increasing
+// handlers due to queue overflow or backpressure. This counter is monotonically
-// and is safe to call concurrently.
+// increasing and is safe to call concurrently.
 func (s *Streamer) GetDroppedMessages() uint64 {
-	return atomic.LoadUint64(&s.totalDropped)
+	return s.totalDropped.Load()
+}
+
+// runHandlerWorker processes messages for a specific handler
+func (s *Streamer) runHandlerWorker(info *handlerInfo) {
+	defer s.workers.Done()
+
+	for msg := range info.queue {
+		start := time.Now()
+
+		info.handler.HandleMessage(msg)
+
+		elapsed := time.Since(start)
+
+		// Update metrics
+		info.metrics.mu.Lock()
+		info.metrics.processedCount++
+		info.metrics.totalTime += elapsed
+
+		// Update min time
+		if elapsed < info.metrics.minTime {
+			info.metrics.minTime = elapsed
+		}
+
+		// Update max time
+		if elapsed > info.metrics.maxTime {
+			info.metrics.maxTime = elapsed
+		}
+		info.metrics.mu.Unlock()
+	}
 }
 
 // logMetrics logs the current streaming statistics
@@ -343,7 +440,8 @@ func (s *Streamer) logMetrics() {
 	uptime := time.Since(metrics.ConnectedSince)
 
 	const bitsPerMegabit = 1000000
-	totalDropped := atomic.LoadUint64(&s.totalDropped)
+
+	totalDropped := s.totalDropped.Load()
 
 	s.logger.Info(
 		"Stream statistics",
@@ -369,16 +467,19 @@ func (s *Streamer) logMetrics() {
 
 	// Log per-handler statistics
 	s.mu.RLock()
+
 	for i, info := range s.handlers {
 		info.metrics.mu.Lock()
 		if info.metrics.processedCount > 0 {
 			// Safe conversion: processedCount is bounded by maxInt64
 			processedCount := info.metrics.processedCount
+
 			const maxInt64 = 1<<63 - 1
+
 			if processedCount > maxInt64 {
 				processedCount = maxInt64
 			}
-			//nolint:gosec // processedCount is explicitly bounded above
+
 			avgTime := info.metrics.totalTime / time.Duration(processedCount)
 			s.logger.Info(
 				"Handler statistics",
@@ -387,7 +488,7 @@ func (s *Streamer) logMetrics() {
 				"queue_len", len(info.queue),
 				"queue_cap", cap(info.queue),
 				"processed", info.metrics.processedCount,
-				"dropped", info.metrics.droppedCount,
+				"dropped", info.metrics.droppedCount.Load(),
 				"avg_time", avgTime,
 				"min_time", info.metrics.minTime,
 				"max_time", info.metrics.maxTime,
@@ -395,6 +496,7 @@ func (s *Streamer) logMetrics() {
 		}
 		info.metrics.mu.Unlock()
 	}
+
 	s.mu.RUnlock()
 }
 
@@ -403,9 +505,10 @@ func (s *Streamer) updateMetrics(messageBytes int) {
 	s.metrics.RecordMessage(int64(messageBytes))
 }
 
-// streamWithReconnect handles streaming with automatic reconnection and exponential backoff
+// streamWithReconnect handles streaming with automatic reconnection and
+// exponential backoff
 func (s *Streamer) streamWithReconnect(ctx context.Context) {
-	backoffDelay := minBackoffDelay
+	backoffDelay := s.minBackoff
 	consecutiveFailures := 0
 
 	for {
@@ -422,8 +525,9 @@ func (s *Streamer) streamWithReconnect(ctx context.Context) {
 		err := s.stream(ctx)
 		streamDuration := time.Since(startTime)
 
-		if err == nil {
+		// Only Stop ends the streamer. A stream that ends any other way,
-			// Clean exit (context cancelled)
+		// with or without an error, is a disconnect and is reconnected.
+		if ctx.Err() != nil {
 			return
 		}
 
@@ -434,17 +538,14 @@ func (s *Streamer) streamWithReconnect(ctx context.Context) {
 			"stream_duration", streamDuration)
 		s.metrics.SetConnected(false)
 
-		// Check if context is cancelled
+		// A stream that lasted more than 30 seconds resets the backoff. That
-		if ctx.Err() != nil {
+		// includes a stream dropped after streamIdleTimeout with no line, so
-			return
+		// a silent feed is retried after the first backoff delay every time.
-		}
-
-		// If we streamed for more than 30 seconds, reset the backoff
-		// This indicates we had a successful connection that received data
 		if streamDuration > 30*time.Second {
 			s.logger.Info("Resetting backoff delay due to successful connection",
 				"stream_duration", streamDuration)
-			backoffDelay = minBackoffDelay
+
+			backoffDelay = s.minBackoff
 			consecutiveFailures = 0
 		} else {
 			// Increment consecutive failures
@@ -477,7 +578,13 @@ func (s *Streamer) stream(ctx context.Context) error {
 	connCtx, connCancel := context.WithCancel(ctx)
 	defer connCancel()
 
-	req, err := http.NewRequestWithContext(ctx, "GET", s.url, nil)
+	// The idle timer drops the connection, by cancelling connCtx, when no
+	// line arrives for s.idleTimeout. Each line read starts it over. It
+	// also covers the wait for the server to answer.
+	idleTimer := time.AfterFunc(s.idleTimeout, connCancel)
+	defer idleTimer.Stop()
+
+	req, err := http.NewRequestWithContext(connCtx, http.MethodGet, s.url, nil)
 	if err != nil {
 		return fmt.Errorf("failed to create request: %w", err)
 	}
@@ -490,13 +597,14 @@ func (s *Streamer) stream(ctx context.Context) error {
 		return fmt.Errorf("failed to connect to RIS Live: %w", err)
 	}
 	defer func() {
-		if err := resp.Body.Close(); err != nil {
+		err := resp.Body.Close()
+		if err != nil {
 			s.logger.Error("Failed to close response body", "error", err)
 		}
 	}()
 
 	if resp.StatusCode != http.StatusOK {
-		return fmt.Errorf("unexpected status code: %d", resp.StatusCode)
+		return fmt.Errorf("%w: %d", errUnexpectedStatus, resp.StatusCode)
 	}
 
 	// Wrap body with counting reader to track actual wire bytes
@@ -510,7 +618,9 @@ func (s *Streamer) stream(ctx context.Context) error {
 			return fmt.Errorf("failed to create gzip reader: %w", err)
 		}
 		defer func() { _ = gzReader.Close() }()
+
 		reader = gzReader
+
 		s.logger.Info("Connected to RIS Live stream", "compression", "gzip")
 	} else {
 		s.logger.Info("Connected to RIS Live stream", "compression", "none")
@@ -518,9 +628,6 @@ func (s *Streamer) stream(ctx context.Context) error {
 
 	s.metrics.SetConnected(true)
 
-	// Track wire bytes for metrics updates
-	var lastWireBytes int64
-
 	// Start metrics logging goroutine
 	metricsTicker := time.NewTicker(metricsLogInterval)
 	defer metricsTicker.Stop()
@@ -537,28 +644,50 @@ func (s *Streamer) stream(ctx context.Context) error {
 	}()
 
 	// Wire byte update ticker - update metrics with actual wire bytes periodically
+	go s.recordWireBytes(connCtx, wireCounter)
+
+	return s.readStream(ctx, connCtx, reader, idleTimer)
+}
+
+// recordWireBytes adds the bytes wireCounter has counted to the metrics once a
+// second until ctx is cancelled.
+func (s *Streamer) recordWireBytes(
+	ctx context.Context, wireCounter *countingReader,
+) {
 	wireUpdateTicker := time.NewTicker(time.Second)
 	defer wireUpdateTicker.Stop()
 
-	go func() {
+	var lastWireBytes int64
+
 	for {
 		select {
 		case <-wireUpdateTicker.C:
 			currentBytes := wireCounter.Count()
+
 			delta := currentBytes - lastWireBytes
 			if delta > 0 {
 				s.metrics.RecordWireBytes(delta)
+
 				lastWireBytes = currentBytes
 			}
-			case <-connCtx.Done():
+		case <-ctx.Done():
 			return
 		}
 	}
-	}()
+}
 
+// readStream hands each line of the stream to processLine until the stream
+// ends, a line fails, or ctx is cancelled. Each line restarts idleTimer, which
+// cancels connCtx when it fires.
+func (s *Streamer) readStream(
+	ctx, connCtx context.Context, reader io.Reader, idleTimer *time.Timer,
+) error {
 	scanner := bufio.NewScanner(reader)
+	scanner.Buffer(nil, maxLineSize)
 
 	for scanner.Scan() {
+		idleTimer.Reset(s.idleTimeout)
+
 		select {
 		case <-ctx.Done():
 			s.logger.Info("Stream stopped by context")
@@ -572,6 +701,32 @@ func (s *Streamer) stream(ctx context.Context) error {
 			continue
 		}
 
+		err := s.processLine(ctx, line)
+		if err != nil {
+			return err
+		}
+	}
+
+	if ctx.Err() != nil {
+		return ctx.Err()
+	}
+
+	if connCtx.Err() != nil {
+		return fmt.Errorf("%w for %s", errStreamIdle, s.idleTimeout)
+	}
+
+	err := scanner.Err()
+	if err != nil {
+		return fmt.Errorf("scanner error: %w", err)
+	}
+
+	return errStreamEnded
+}
+
+// processLine passes one line of the stream to the raw handler, parses it,
+// updates the peer set (updatePeers), and dispatches an UPDATE or STATE
+// message to the handlers. An error ends the connection.
+func (s *Streamer) processLine(ctx context.Context, line []byte) error {
 	// Update metrics with decompressed message size
 	s.updateMetrics(len(line))
 
@@ -587,7 +742,9 @@ func (s *Streamer) stream(ctx context.Context) error {
 
 	// Parse the message first
 	var wrapper ristypes.RISLiveMessage
-		if err := json.Unmarshal(line, &wrapper); err != nil {
+
+	err := json.Unmarshal(line, &wrapper)
+	if err != nil {
 		// Log the error and return to trigger reconnection
 		s.logger.Error("Failed to parse JSON",
 			"error", err,
@@ -604,7 +761,7 @@ func (s *Streamer) stream(ctx context.Context) error {
 			"line", string(line),
 		)
 
-			continue
+		return nil
 	}
 
 	// Get the actual message
@@ -613,84 +770,100 @@ func (s *Streamer) stream(ctx context.Context) error {
 	// Parse the timestamp
 	msg.ParsedTimestamp = time.Unix(int64(msg.Timestamp), 0).UTC()
 
+	peerCount := s.updatePeers(&msg)
+
 	// Process based on message type
 	switch msg.Type {
-		case "UPDATE":
+	case ristypes.MessageTypeUpdate:
 		// Process BGP UPDATE messages
 		// Will be dispatched to handlers
 	case "RIS_PEER_STATE":
 		// RIS peer state messages - silently ignore
-			continue
+		return nil
 	case "KEEPALIVE":
 		// BGP keepalive messages - silently process
-			continue
+		return nil
 	case "OPEN":
-			// BGP open messages - track peer as active
-			s.bgpPeersMu.Lock()
-			s.bgpPeers[msg.Peer] = true
-			peerCount := len(s.bgpPeers)
-			s.bgpPeersMu.Unlock()
-			s.metrics.SetBGPPeerCount(peerCount)
-
 		s.logger.Info("BGP session opened",
 			"peer", msg.Peer,
 			"peer_asn", msg.PeerASN,
 			"total_peers", peerCount,
 		)
 
-			continue
+		return nil
-		case "NOTIFICATION":
+	case messageTypeNotification:
-			// BGP notification messages (session closed)
-			s.bgpPeersMu.Lock()
-			delete(s.bgpPeers, msg.Peer)
-			peerCount := len(s.bgpPeers)
-			s.bgpPeersMu.Unlock()
-			s.metrics.SetBGPPeerCount(peerCount)
-
 		s.logger.Warn("BGP notification",
 			"peer", msg.Peer,
 			"peer_asn", msg.PeerASN,
 			"total_peers", peerCount,
 		)
 
-			continue
+		return nil
-		case "STATE":
+	case ristypes.MessageTypeState:
-			// Peer state changes - silently ignore
+		// The BGP session with the peer changed state, for example went
-			continue
+		// down. Dispatched to the handlers that want it.
 	default:
 		s.logger.Warn("Unknown message type, skipping",
 			"type", msg.Type,
 		)
 
-			continue
+		return nil
 	}
 
-		// Dispatch to interested handlers. Stop cancels ctx and closes the
+	return s.dispatchMessage(ctx, &msg)
-		// queues under the write lock, so if ctx is cancelled here, under the
+}
-		// read lock, the queues are closed and must not be sent to.
+
+// updatePeers adds the peer msg came from to the peer set, or removes it if
+// msg ends the peer's session: a NOTIFICATION, or a STATE message whose state
+// is down. It records the size of the set as the BGP peer count and returns
+// it.
+func (s *Streamer) updatePeers(msg *ristypes.RISMessage) int {
+	sessionEnded := msg.Type == messageTypeNotification ||
+		(msg.Type == ristypes.MessageTypeState && msg.State == ristypes.StateDown)
+
+	s.bgpPeersMu.Lock()
+
+	if sessionEnded {
+		delete(s.bgpPeers, msg.Peer)
+	} else {
+		s.bgpPeers[msg.Peer] = true
+	}
+
+	peerCount := len(s.bgpPeers)
+	s.bgpPeersMu.Unlock()
+	s.metrics.SetBGPPeerCount(peerCount)
+
+	return peerCount
+}
+
+// dispatchMessage queues msg for each handler that wants it, unless
+// queueMessage drops it for that handler.
+func (s *Streamer) dispatchMessage(
+	ctx context.Context, msg *ristypes.RISMessage,
+) error {
+	// Stop cancels ctx and closes the queues under the write lock, so if ctx
+	// is cancelled here, under the read lock, the queues are closed and must
+	// not be sent to.
 	s.mu.RLock()
+
 	if ctx.Err() != nil {
 		s.mu.RUnlock()
 
 		return ctx.Err()
 	}
+
 	for _, info := range s.handlers {
 		if !info.handler.WantsMessage(msg.Type) {
 			continue
 		}
 
-			// Check if we should drop due to backpressure
+		if !s.queueMessage(info, msg) {
-			if s.shouldDropForBackpressure(info) {
+			info.metrics.droppedCount.Add(1)
-				atomic.AddUint64(&info.metrics.droppedCount, 1)
+			s.totalDropped.Add(1)
-				atomic.AddUint64(&s.totalDropped, 1)
 
 			continue
 		}
 
-			// Try to queue the message
-			select {
-			case info.queue <- &msg:
-				// Message queued successfully
 		// Update high water mark if needed
 		queueLen := len(info.queue)
 		info.metrics.mu.Lock()
@@ -698,23 +871,61 @@ func (s *Streamer) stream(ctx context.Context) error {
 			info.metrics.queueHighWaterMark = queueLen
 		}
 		info.metrics.mu.Unlock()
-			default:
-				// Queue is full, drop the message
-				atomic.AddUint64(&info.metrics.droppedCount, 1)
-				atomic.AddUint64(&s.totalDropped, 1)
-			}
-		}
-		s.mu.RUnlock()
 	}
 
-	if err := scanner.Err(); err != nil {
+	s.mu.RUnlock()
-		return fmt.Errorf("scanner error: %w", err)
-	}
 
 	return nil
 }
 
-// shouldDropForBackpressure determines if a message should be dropped based on queue utilization
+// queueMessage puts msg on the handler's queue and reports false if it dropped
+// msg instead. It never waits for room, so a full queue does not hold up the
+// read loop.
+//
+// A message is dropped while the queue is filling up (shouldDropForBackpressure)
+// or full, except a STATE message: one whose state is down is the only notice
+// that all of a peer's routes are gone. A STATE message skips the random drop,
+// and one that finds the queue full is held in info.heldState. Held messages
+// go into the queue, oldest first, before anything else for the handler, as
+// soon as there is room, or when Stop is called; while any is still held, the
+// handler's other messages are dropped, so none of them overtakes a held down.
+//
+// The held list stays small: a STATE message is a few hundred bytes and the
+// feed sends about ten a second, so a queue full for a minute holds about 600.
+func (s *Streamer) queueMessage(
+	info *handlerInfo, msg *ristypes.RISMessage,
+) bool {
+	for len(info.heldState) > 0 && trySend(info.queue, info.heldState[0]) {
+		info.heldState = info.heldState[1:]
+	}
+
+	if msg.Type == ristypes.MessageTypeState {
+		if len(info.heldState) > 0 || !trySend(info.queue, msg) {
+			info.heldState = append(info.heldState, msg)
+		}
+
+		return true
+	}
+
+	if len(info.heldState) > 0 || s.shouldDropForBackpressure(info) {
+		return false
+	}
+
+	return trySend(info.queue, msg)
+}
+
+// trySend puts msg on queue if there is room and reports whether it did.
+func trySend(queue chan *ristypes.RISMessage, msg *ristypes.RISMessage) bool {
+	select {
+	case queue <- msg:
+		return true
+	default:
+		return false
+	}
+}
+
+// shouldDropForBackpressure determines if a message should be dropped based on
+// queue utilization
 func (s *Streamer) shouldDropForBackpressure(info *handlerInfo) bool {
 	// Calculate queue utilization
 	queueLen := len(info.queue)

internal/streamer/streamer_internal_test.go

@@ -0,0 +1,850 @@
+package streamer
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"errors"
+	"io"
+	"log/slog"
+	"net/http"
+	"net/http/httptest"
+	"runtime"
+	"strings"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"git.eeqj.de/sneak/routewatch/internal/logger"
+	"git.eeqj.de/sneak/routewatch/internal/metrics"
+	"git.eeqj.de/sneak/routewatch/internal/ristypes"
+)
+
+func TestNewStreamer(t *testing.T) {
+	t.Parallel()
+
+	logger := logger.New()
+	metricsTracker := metrics.New()
+	s := New(logger, metricsTracker)
+
+	if s == nil {
+		t.Fatal("New() returned nil")
+	}
+
+	if s.logger != logger {
+		t.Error("logger not set correctly")
+	}
+
+	if s.client == nil {
+		t.Error("HTTP client not initialized")
+	}
+
+	if s.handlers == nil {
+		t.Error("handlers slice not initialized")
+	}
+
+	if s.metrics != metricsTracker {
+		t.Error("metrics tracker not set correctly")
+	}
+
+	if s.idleTimeout != streamIdleTimeout {
+		t.Errorf("idleTimeout = %s, want %s", s.idleTimeout, streamIdleTimeout)
+	}
+
+	if s.minBackoff != minBackoffDelay {
+		t.Errorf("minBackoff = %s, want %s", s.minBackoff, minBackoffDelay)
+	}
+}
+
+// TestStreamDoesNotLeakTickersAcrossReconnects drives many short-lived
+// connections (each stream call is one reconnect cycle) and asserts the
+// goroutine count returns to its starting value. Each connection starts two
+// ticker goroutines; before the fix they lived until the streamer's lifetime
+// context was cancelled, so every reconnect leaked two.
+//
+//nolint:paralleltest // counts goroutines, so no other test may run alongside
+func TestStreamDoesNotLeakTickersAcrossReconnects(t *testing.T) {
+	// The handler returns immediately, so the response body is empty and each
+	// stream call ends at once, standing in for a dropped connection.
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(_ http.ResponseWriter, _ *http.Request) {}))
+	defer srv.Close()
+
+	s := New(logger.New(), metrics.New())
+	s.url = srv.URL
+
+	// One warm-up connection so any persistent HTTP transport goroutine exists
+	// before we take the baseline.
+	err := s.stream(context.Background())
+	if !errors.Is(err, errStreamEnded) {
+		t.Fatalf("warm-up stream returned %v, want %v", err, errStreamEnded)
+	}
+
+	s.client.CloseIdleConnections()
+
+	baseline := settledGoroutineCount()
+
+	const reconnects = 20
+	for range reconnects {
+		err := s.stream(context.Background())
+		if !errors.Is(err, errStreamEnded) {
+			t.Fatalf("stream returned %v, want %v", err, errStreamEnded)
+		}
+	}
+
+	s.client.CloseIdleConnections()
+
+	if !waitForGoroutines(baseline) {
+		t.Fatalf("goroutines did not return to baseline %d after %d reconnects, got %d",
+			baseline, reconnects, runtime.NumGoroutine())
+	}
+}
+
+// updateLine is one UPDATE message as RIS Live sends it.
+const updateLine = `{"type":"ris_message","data":{"type":"UPDATE",` +
+	`"peer":"192.0.2.1","peer_asn":"64496","timestamp":1700000000}}` + "\n"
+
+// updateHandler wants UPDATE messages. For each one it handles it sends on
+// handled if there is room, and otherwise does nothing; a nil handled is
+// never sent on.
+type updateHandler struct {
+	handled chan struct{}
+}
+
+func (updateHandler) WantsMessage(messageType string) bool {
+	return messageType == "UPDATE"
+}
+
+func (updateHandler) QueueCapacity() int { return 10 }
+
+func (h updateHandler) HandleMessage(*ristypes.RISMessage) {
+	select {
+	case h.handled <- struct{}{}:
+	default:
+	}
+}
+
+// flush sends what w has buffered to the client. The writers httptest servers
+// hand to handlers can always do that.
+func flush(w http.ResponseWriter) {
+	flusher, ok := w.(http.Flusher)
+	if ok {
+		flusher.Flush()
+	}
+}
+
+// stateDownLine is the STATE message RIS Live sends when the collector's BGP
+// session with the peer stateDownPeer goes down.
+const (
+	stateDownPeer = "192.0.2.1"
+	stateDownLine = `{"type":"ris_message","data":{"type":"STATE",` +
+		`"peer":"` + stateDownPeer + `",` +
+		`"peer_asn":"64496","timestamp":1700000000,"state":"down"}}` + "\n"
+)
+
+// stateHandler wants STATE messages only. Tests read what it is handed from
+// its queue, so HandleMessage does nothing.
+type stateHandler struct{}
+
+func (stateHandler) WantsMessage(messageType string) bool {
+	return messageType == ristypes.MessageTypeState
+}
+
+func (stateHandler) QueueCapacity() int { return 10 }
+
+func (stateHandler) HandleMessage(*ristypes.RISMessage) {}
+
+// TestStreamHandsStateMessagesToHandlersThatWantThem has the server send one
+// STATE message whose state is down. The streamer must queue it, with its
+// peer and state, for a handler that wants STATE messages.
+func TestStreamHandsStateMessagesToHandlersThatWantThem(t *testing.T) {
+	t.Parallel()
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			_, _ = io.WriteString(w, stateDownLine)
+		}))
+	defer srv.Close()
+
+	s := New(logger.New(), metrics.New())
+	s.url = srv.URL
+	s.RegisterHandler(stateHandler{})
+
+	err := s.stream(context.Background())
+	if !errors.Is(err, errStreamEnded) {
+		t.Fatalf("stream returned %v, want %v", err, errStreamEnded)
+	}
+
+	select {
+	case msg := <-s.handlers[0].queue:
+		if msg.Type != ristypes.MessageTypeState || msg.Peer != stateDownPeer ||
+			msg.State != ristypes.StateDown {
+			t.Errorf("queued message has type %q, peer %q, state %q; "+
+				"want %q, %q, %q", msg.Type, msg.Peer, msg.State,
+				ristypes.MessageTypeState, stateDownPeer, ristypes.StateDown)
+		}
+	default:
+		t.Fatal("the STATE message was not queued for the handler that wants it")
+	}
+}
+
+// TestHeldStateMessagesKeepTheirPlaceInTheQueue hands queueMessage two STATE
+// messages and two updates while the queue is full, making room in between.
+// Neither STATE message may be dropped: each must reach the queue after
+// everything queued before it and before anything handed over after it. The
+// update handed over while a STATE message is held is dropped.
+func TestHeldStateMessagesKeepTheirPlaceInTheQueue(t *testing.T) {
+	t.Parallel()
+
+	s := New(logger.New(), metrics.New())
+	s.RegisterHandler(stateHandler{})
+
+	info := s.handlers[0]
+
+	before := &ristypes.RISMessage{Type: ristypes.MessageTypeUpdate}
+	for len(info.queue) < cap(info.queue) {
+		info.queue <- before
+	}
+
+	firstDown := &ristypes.RISMessage{
+		Type: ristypes.MessageTypeState, State: ristypes.StateDown,
+	}
+	secondDown := &ristypes.RISMessage{
+		Type: ristypes.MessageTypeState, State: ristypes.StateDown,
+	}
+	updateWhileHeld := &ristypes.RISMessage{Type: ristypes.MessageTypeUpdate}
+	after := &ristypes.RISMessage{Type: ristypes.MessageTypeUpdate}
+
+	if !s.queueMessage(info, firstDown) {
+		t.Fatal("the first STATE message was dropped on a full queue")
+	}
+
+	<-info.queue // the handler makes room for one message
+
+	if !s.queueMessage(info, secondDown) {
+		t.Fatal("the second STATE message was dropped")
+	}
+
+	if s.queueMessage(info, updateWhileHeld) {
+		t.Fatal("an update was queued while a STATE message was held")
+	}
+
+	// The handler takes every message queued before the first STATE message.
+	for range cap(info.queue) - 1 {
+		if <-info.queue != before {
+			t.Fatal("a message overtook those queued before the first STATE message")
+		}
+	}
+
+	if !s.queueMessage(info, after) {
+		t.Fatal("an update was dropped on a queue with room")
+	}
+
+	if <-info.queue != firstDown {
+		t.Fatal("the first STATE message did not come next")
+	}
+
+	if <-info.queue != secondDown {
+		t.Fatal("the second STATE message did not come after the first")
+	}
+
+	if <-info.queue != after {
+		t.Fatal("the update handed over after the STATE messages did not come " +
+			"after them")
+	}
+
+	if len(info.queue) != 0 {
+		t.Errorf("%d messages left on the queue, want none", len(info.queue))
+	}
+}
+
+// takeAll takes the messages off queue, as a handler's worker would, and sends
+// them on the returned channel once queue is closed.
+func takeAll(queue chan *ristypes.RISMessage) chan []*ristypes.RISMessage {
+	taken := make(chan []*ristypes.RISMessage, 1)
+
+	go func() {
+		var msgs []*ristypes.RISMessage
+
+		for msg := range queue {
+			msgs = append(msgs, msg)
+		}
+
+		taken <- msgs
+	}()
+
+	return taken
+}
+
+// newStreamerWithFullStateQueue returns a streamer reading from url with one
+// handler, a stateHandler whose queue is full.
+func newStreamerWithFullStateQueue(url string) *Streamer {
+	s := New(logger.New(), metrics.New())
+	s.url = url
+	s.RegisterHandler(stateHandler{})
+
+	queue := s.handlers[0].queue
+	for len(queue) < cap(queue) {
+		queue <- &ristypes.RISMessage{Type: ristypes.MessageTypeState}
+	}
+
+	return s
+}
+
+// TestOtherHandlersKeepReceivingWhileAStateMessageIsHeld has the server send a
+// STATE message whose state is down, then an UPDATE, while the queue of the
+// handler that wants STATE messages is full. The STATE message must be held
+// for that handler, and the read loop must go on to queue the UPDATE for the
+// handler that wants it.
+func TestOtherHandlersKeepReceivingWhileAStateMessageIsHeld(t *testing.T) {
+	t.Parallel()
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			_, _ = io.WriteString(w, stateDownLine+updateLine)
+		}))
+	defer srv.Close()
+
+	s := newStreamerWithFullStateQueue(srv.URL)
+	s.RegisterHandler(updateHandler{})
+
+	err := s.stream(context.Background())
+	if !errors.Is(err, errStreamEnded) {
+		t.Fatalf("stream returned %v, want %v", err, errStreamEnded)
+	}
+
+	held := s.handlers[0].heldState
+	if len(held) != 1 || held[0].Peer != stateDownPeer ||
+		held[0].State != ristypes.StateDown {
+		t.Errorf("held %d STATE messages, want the down from %s alone",
+			len(held), stateDownPeer)
+	}
+
+	got := len(s.handlers[1].queue)
+	if got != 1 {
+		t.Errorf("%d messages queued for the handler that wants UPDATE "+
+			"messages, want 1", got)
+	}
+}
+
+// TestStopQueuesTheHeldStateMessages has the server send two STATE messages
+// and an UPDATE and hold the response open while the queue of the handler that
+// wants STATE messages is full, so both are held. Stop must put both on the
+// queue, after the messages already on it, before it closes the queue, and
+// must end the stream.
+func TestStopQueuesTheHeldStateMessages(t *testing.T) {
+	t.Parallel()
+
+	const (
+		held    = 2
+		lines   = held + 1
+		maxWait = 5 * time.Second
+	)
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, r *http.Request) {
+			_, _ = io.WriteString(w, stateDownLine+stateDownLine+updateLine)
+
+			flush(w)
+			<-r.Context().Done()
+		}))
+	defer srv.Close()
+
+	s := newStreamerWithFullStateQueue(srv.URL)
+	// The raw handler sees each line before the read loop hands it to the
+	// queues, so once it has seen the UPDATE, both STATE messages are held.
+	linesSeen := make(chan struct{}, lines)
+
+	s.RegisterRawHandler(func(string) { linesSeen <- struct{}{} })
+
+	// As in TestStopBeforeMessageReachesQueues, setting cancel as Start does
+	// lets Stop cancel the stream run here.
+	ctx, cancel := context.WithCancel(context.Background())
+	s.cancel = cancel
+	streamErr := make(chan error, 1)
+
+	go func() { streamErr <- s.stream(ctx) }()
+
+	for range lines {
+		select {
+		case <-linesSeen:
+		case <-time.After(maxWait):
+			t.Fatalf("the read loop did not read %d lines within %s",
+				lines, maxWait)
+		}
+	}
+
+	// No worker runs here: the test takes from the full queue in its place.
+	queue := s.handlers[0].queue
+	taken := takeAll(queue)
+	stopped := make(chan struct{})
+
+	go func() {
+		s.Stop(t.Context())
+		close(stopped)
+	}()
+
+	select {
+	case <-stopped:
+	case <-time.After(maxWait):
+		t.Fatalf("Stop did not return within %s while a STATE message was held",
+			maxWait)
+	}
+
+	// Stop has closed the queue, so takeAll is done or about to be.
+	msgs := <-taken
+	if len(msgs) != cap(queue)+held {
+		t.Fatalf("%d messages came off the queue, want the %d on it and the "+
+			"%d held", len(msgs), cap(queue), held)
+	}
+
+	for _, msg := range msgs[cap(queue):] {
+		if msg.Peer != stateDownPeer || msg.State != ristypes.StateDown {
+			t.Errorf("message after those on the queue has peer %q, state %q; "+
+				"want the down from %s", msg.Peer, msg.State, stateDownPeer)
+		}
+	}
+
+	select {
+	case err := <-streamErr:
+		if !errors.Is(err, context.Canceled) {
+			t.Errorf("stream returned %v, want %v", err, context.Canceled)
+		}
+	case <-time.After(maxWait):
+		t.Fatalf("stream did not end within %s of Stop", maxWait)
+	}
+}
+
+// TestStopReturnsByItsDeadline stops a streamer whose handler has a full queue
+// and a STATE message held for it, and a worker that stays busy until the test
+// ends, as one does whose handler waits for a database write. Stop must return
+// within its deadline plus a small margin instead of waiting for room on the
+// queue and for the worker, and log as not handed over the held message and
+// every message on the queue.
+func TestStopReturnsByItsDeadline(t *testing.T) {
+	t.Parallel()
+
+	const (
+		deadline = 100 * time.Millisecond
+		margin   = 2 * time.Second
+	)
+
+	var logs bytes.Buffer
+
+	s := newStreamerWithFullStateQueue("")
+	s.logger = &logger.Logger{Logger: slog.New(slog.NewJSONHandler(&logs,
+		&slog.HandlerOptions{Level: slog.LevelError}))}
+	s.handlers[0].heldState = []*ristypes.RISMessage{
+		{Type: ristypes.MessageTypeState},
+	}
+	wantNotHandedOver := 1 + len(s.handlers[0].queue)
+	// Stop cancels the stream Start would have started; there is none here.
+	s.cancel = func() {}
+
+	release := make(chan struct{})
+	defer close(release)
+
+	s.workers.Go(func() { <-release })
+
+	ctx, cancel := context.WithTimeout(t.Context(), deadline)
+	defer cancel()
+
+	stopped := make(chan struct{})
+
+	go func() {
+		s.Stop(ctx)
+		close(stopped)
+	}()
+
+	select {
+	case <-stopped:
+	case <-time.After(deadline + margin):
+		t.Fatalf("Stop did not return within %s of its deadline", margin)
+	}
+
+	var logged struct {
+		Messages int `json:"messages"`
+	}
+
+	err := json.Unmarshal(logs.Bytes(), &logged)
+	if err != nil {
+		t.Fatalf("want the error Stop logs, got %q: %v", logs.String(), err)
+	}
+
+	if logged.Messages != wantNotHandedOver {
+		t.Errorf("Stop logged %d messages not handed over, want %d",
+			logged.Messages, wantNotHandedOver)
+	}
+}
+
+// TestStopBeforeMessageReachesQueues stops the streamer after the read loop
+// has checked for cancellation but before it hands the message to the handler
+// queues. That is the gap a stop from another goroutine can land in, and it
+// used to end in "send on closed channel". The raw handler runs in that gap on
+// the read loop itself, so calling Stop from it hits the gap every time.
+func TestStopBeforeMessageReachesQueues(t *testing.T) {
+	t.Parallel()
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			_, _ = io.WriteString(w, updateLine)
+		}))
+	defer srv.Close()
+
+	s := New(logger.New(), metrics.New())
+	s.url = srv.URL
+	s.RegisterHandler(updateHandler{})
+	s.RegisterRawHandler(func(string) { s.Stop(t.Context()) })
+
+	// Start would run the stream in the background, where the test cannot
+	// wait for it. Setting cancel as Start does lets Stop cancel the stream
+	// run here instead.
+	ctx, cancel := context.WithCancel(context.Background())
+	s.cancel = cancel
+
+	err := s.stream(ctx)
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("stream returned %v, want %v", err, context.Canceled)
+	}
+
+	// A second Stop must not close the queues again.
+	s.Stop(t.Context())
+}
+
+// TestReconnectWhenStreamEnds has the server end the first response
+// normally, as when RIS Live restarts. The streamer must treat that as a
+// disconnect and connect again.
+func TestReconnectWhenStreamEnds(t *testing.T) {
+	t.Parallel()
+
+	endResponse := func(http.ResponseWriter, *http.Request) {}
+	checkReconnect(t, endResponse, streamIdleTimeout)
+}
+
+// TestReconnectWhenStreamGoesIdle has the server hold the first response
+// open without sending anything. The streamer must drop it once the idle
+// timeout passes and connect again.
+func TestReconnectWhenStreamGoesIdle(t *testing.T) {
+	t.Parallel()
+
+	const idleTimeout = 50 * time.Millisecond
+
+	holdResponse := func(w http.ResponseWriter, r *http.Request) {
+		flush(w)
+		<-r.Context().Done()
+	}
+	checkReconnect(t, holdResponse, idleTimeout)
+}
+
+// TestReconnectWhenServerNeverAnswers has the server accept the first
+// request and never answer it, not even with headers. The streamer must give
+// up on it once the idle timeout passes and connect again.
+func TestReconnectWhenServerNeverAnswers(t *testing.T) {
+	t.Parallel()
+
+	const idleTimeout = 50 * time.Millisecond
+
+	neverAnswer := func(_ http.ResponseWriter, r *http.Request) {
+		<-r.Context().Done()
+	}
+	checkReconnect(t, neverAnswer, idleTimeout)
+}
+
+// TestStreamNotDroppedWhileLinesArrive has the server send lines well within
+// the idle timeout for several times the timeout, then end the response. Each
+// line must start the idle timeout over, so the stream lasts until the server
+// ends it instead of being dropped as idle.
+func TestStreamNotDroppedWhileLinesArrive(t *testing.T) {
+	t.Parallel()
+
+	const (
+		idleTimeout  = 100 * time.Millisecond
+		lineInterval = idleTimeout / 10
+		sendFor      = 5 * idleTimeout
+	)
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			deadline := time.Now().Add(sendFor)
+			for time.Now().Before(deadline) {
+				_, _ = io.WriteString(w, updateLine)
+
+				flush(w)
+				time.Sleep(lineInterval)
+			}
+		}))
+	defer srv.Close()
+
+	s := New(logger.New(), metrics.New())
+	s.url = srv.URL
+	s.idleTimeout = idleTimeout
+
+	err := s.stream(context.Background())
+	if !errors.Is(err, errStreamEnded) {
+		t.Fatalf("stream returned %v, want %v", err, errStreamEnded)
+	}
+}
+
+// TestStreamReadsLongLine has the server send one UPDATE line of about
+// 200 KiB, padded with raw hex the way RIS Live writes an extended BGP
+// message, and then end the response. The line is longer than a
+// bufio.Scanner reads by default. The streamer must hand the message to the
+// handler's queue and read on until the server ends the stream, instead of
+// dropping the connection at the long line.
+func TestStreamReadsLongLine(t *testing.T) {
+	t.Parallel()
+
+	const rawHexLength = 200 * bytesPerKB
+
+	longLine := `{"type":"ris_message","data":{"type":"UPDATE","peer":"192.0.2.1",` +
+		`"peer_asn":"64496","timestamp":1700000000,"raw":"` +
+		strings.Repeat("0", rawHexLength) + `"}}` + "\n"
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			_, _ = io.WriteString(w, longLine)
+		}))
+	defer srv.Close()
+
+	s := New(logger.New(), metrics.New())
+	s.url = srv.URL
+	s.RegisterHandler(updateHandler{})
+
+	err := s.stream(context.Background())
+	if !errors.Is(err, errStreamEnded) {
+		t.Fatalf("stream returned %v, want %v", err, errStreamEnded)
+	}
+
+	queued := len(s.handlers[0].queue)
+	if queued != 1 {
+		t.Errorf("handler queue holds %d messages, want 1", queued)
+	}
+}
+
+// blockedHandler wants UPDATE messages, has room on its queue for one, and
+// does not return from HandleMessage until release is closed.
+type blockedHandler struct {
+	release chan struct{}
+}
+
+func (blockedHandler) WantsMessage(messageType string) bool {
+	return messageType == ristypes.MessageTypeUpdate
+}
+
+func (blockedHandler) QueueCapacity() int { return 1 }
+
+func (h blockedHandler) HandleMessage(*ristypes.RISMessage) { <-h.release }
+
+// TestHandlerStatsReadWhileMessagesAreDropped has the server send a few UPDATE
+// lines to a blockedHandler, so the read loop drops all but the one the
+// handler holds and the one on its queue, while the test reads the handler's
+// statistics. The drop count must reach the number dropped, and the race
+// detector must find the count read and written safely.
+func TestHandlerStatsReadWhileMessagesAreDropped(t *testing.T) {
+	t.Parallel()
+
+	const (
+		lines       = 5
+		wantDropped = lines - 2
+		maxWait     = 5 * time.Second
+	)
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, r *http.Request) {
+			_, _ = io.WriteString(w, strings.Repeat(updateLine, lines))
+
+			flush(w)
+			<-r.Context().Done()
+		}))
+	defer srv.Close()
+
+	s := New(logger.New(), metrics.New())
+	s.url = srv.URL
+	handler := blockedHandler{release: make(chan struct{})}
+	s.RegisterHandler(handler)
+
+	err := s.Start(t.Context())
+	if err != nil {
+		t.Fatalf("Start returned error: %v", err)
+	}
+	defer s.Stop(t.Context())
+	defer close(handler.release)
+
+	// The test polls instead of waiting for a signal from the read loop: a
+	// signal would order the drops before the reads here, and the race
+	// detector would have nothing to report.
+	deadline := time.Now().Add(maxWait)
+
+	for s.GetHandlerStats()[0].DroppedCount < wantDropped {
+		if time.Now().After(deadline) {
+			t.Fatalf("fewer than %d messages counted as dropped within %s",
+				wantDropped, maxWait)
+		}
+	}
+}
+
+// TestPeerCountCountsEveryPeerHeardFrom has the server send an UPDATE from
+// stateDownPeer and a KEEPALIVE from another peer, with no OPEN from either,
+// and then a message that ends stateDownPeer's session: a NOTIFICATION, or a
+// STATE message whose state is down. The peer count must be 2 after the first
+// two messages and 1 after the third.
+func TestPeerCountCountsEveryPeerHeardFrom(t *testing.T) {
+	t.Parallel()
+
+	const (
+		otherPeer      = "192.0.2.2"
+		peerUpdateLine = `{"type":"ris_message","data":{"type":"UPDATE",` +
+			`"peer":"` + stateDownPeer + `",` +
+			`"peer_asn":"64496","timestamp":1700000000}}` + "\n"
+		otherPeerKeepaliveLine = `{"type":"ris_message","data":{"type":"KEEPALIVE",` +
+			`"peer":"` + otherPeer + `",` +
+			`"peer_asn":"64497","timestamp":1700000000}}` + "\n"
+		peerNotificationLine = `{"type":"ris_message","data":{"type":"NOTIFICATION",` +
+			`"peer":"` + stateDownPeer + `",` +
+			`"peer_asn":"64496","timestamp":1700000000}}` + "\n"
+		bothPeers = 2
+	)
+
+	sessionEndLines := map[string]string{
+		"NOTIFICATION": peerNotificationLine,
+		"STATE down":   stateDownLine,
+	}
+
+	for name, sessionEndLine := range sessionEndLines {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			tracker := metrics.New()
+			s := New(logger.New(), tracker)
+
+			streamLines(t, s, peerUpdateLine+otherPeerKeepaliveLine)
+
+			got := tracker.GetBGPPeerCount()
+			if got != bothPeers {
+				t.Fatalf("peer count is %d after messages from two peers, want %d",
+					got, bothPeers)
+			}
+
+			streamLines(t, s, sessionEndLine)
+
+			got = tracker.GetBGPPeerCount()
+			if got != 1 {
+				t.Errorf("peer count is %d after one of two peers' sessions "+
+					"ended, want 1", got)
+			}
+		})
+	}
+}
+
+// streamLines has a server send lines as its whole response and runs one
+// stream of s against it.
+func streamLines(t *testing.T, s *Streamer, lines string) {
+	t.Helper()
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, _ *http.Request) {
+			_, _ = io.WriteString(w, lines)
+		}))
+	defer srv.Close()
+
+	s.url = srv.URL
+
+	err := s.stream(context.Background())
+	if !errors.Is(err, errStreamEnded) {
+		t.Fatalf("stream returned %v, want %v", err, errStreamEnded)
+	}
+}
+
+// checkReconnect runs a streamer against a fake RIS Live server. The first
+// response is served by firstResponse; every later one carries one UPDATE
+// line and then stays open. The streamer must come back for a second
+// response, must not be marked connected when it does, and must handle the
+// UPDATE in it.
+func checkReconnect(
+	t *testing.T, firstResponse http.HandlerFunc, idleTimeout time.Duration,
+) {
+	t.Helper()
+
+	const maxWait = 5 * time.Second
+
+	tracker := metrics.New()
+	connectedAtReconnect := make(chan bool, 1)
+
+	var requests atomic.Int32
+
+	srv := httptest.NewServer(http.HandlerFunc(
+		func(w http.ResponseWriter, r *http.Request) {
+			if requests.Add(1) == 1 {
+				firstResponse(w, r)
+
+				return
+			}
+
+			select {
+			case connectedAtReconnect <- tracker.IsConnected():
+			default:
+			}
+
+			_, _ = io.WriteString(w, updateLine)
+
+			flush(w)
+			<-r.Context().Done()
+		}))
+	defer srv.Close()
+
+	s := New(logger.New(), tracker)
+	s.url = srv.URL
+	s.idleTimeout = idleTimeout
+	s.minBackoff = time.Millisecond
+	handler := updateHandler{handled: make(chan struct{}, 1)}
+	s.RegisterHandler(handler)
+
+	err := s.Start(t.Context())
+	if err != nil {
+		t.Fatalf("Start returned error: %v", err)
+	}
+	defer s.Stop(t.Context())
+
+	select {
+	case <-handler.handled:
+	case <-time.After(maxWait):
+		t.Fatalf("no message from a second response handled within %s", maxWait)
+	}
+
+	if <-connectedAtReconnect {
+		t.Error("streamer was still marked connected when it came back " +
+			"for a second response")
+	}
+}
+
+// settledGoroutineCount lets transient goroutines finish, then reports the
+// current count.
+func settledGoroutineCount() int {
+	prev := runtime.NumGoroutine()
+
+	for range 20 {
+		time.Sleep(10 * time.Millisecond)
+
+		cur := runtime.NumGoroutine()
+		if cur == prev {
+			return cur
+		}
+
+		prev = cur
+	}
+
+	return prev
+}
+
+// waitForGoroutines waits until the goroutine count drops to target or below.
+func waitForGoroutines(target int) bool {
+	for range 100 {
+		if runtime.NumGoroutine() <= target {
+			return true
+		}
+
+		time.Sleep(10 * time.Millisecond)
+	}
+
+	return false
+}

internal/streamer/streamer_test.go

@@ -1,146 +0,0 @@
-package streamer
-
-import (
-	"context"
-	"errors"
-	"io"
-	"net/http"
-	"net/http/httptest"
-	"runtime"
-	"testing"
-	"time"
-
-	"git.eeqj.de/sneak/routewatch/internal/logger"
-	"git.eeqj.de/sneak/routewatch/internal/metrics"
-	"git.eeqj.de/sneak/routewatch/internal/ristypes"
-)
-
-func TestNewStreamer(t *testing.T) {
-	logger := logger.New()
-	metricsTracker := metrics.New()
-	s := New(logger, metricsTracker)
-
-	if s == nil {
-		t.Fatal("New() returned nil")
-	}
-
-	if s.logger != logger {
-		t.Error("logger not set correctly")
-	}
-
-	if s.client == nil {
-		t.Error("HTTP client not initialized")
-	}
-
-	if s.handlers == nil {
-		t.Error("handlers slice not initialized")
-	}
-
-	if s.metrics != metricsTracker {
-		t.Error("metrics tracker not set correctly")
-	}
-}
-
-// TestStreamDoesNotLeakTickersAcrossReconnects drives many short-lived
-// connections (each stream call is one reconnect cycle) and asserts the
-// goroutine count returns to its starting value. Each connection starts two
-// ticker goroutines; before the fix they lived until the streamer's lifetime
-// context was cancelled, so every reconnect leaked two.
-func TestStreamDoesNotLeakTickersAcrossReconnects(t *testing.T) {
-	// The handler returns immediately, so the response body is empty and each
-	// stream call ends at once, standing in for a dropped connection.
-	srv := httptest.NewServer(http.HandlerFunc(func(_ http.ResponseWriter, _ *http.Request) {}))
-	defer srv.Close()
-
-	s := New(logger.New(), metrics.New())
-	s.url = srv.URL
-
-	// One warm-up connection so any persistent HTTP transport goroutine exists
-	// before we take the baseline.
-	if err := s.stream(context.Background()); err != nil {
-		t.Fatalf("warm-up stream returned error: %v", err)
-	}
-	s.client.CloseIdleConnections()
-
-	baseline := settledGoroutineCount()
-
-	const reconnects = 20
-	for range reconnects {
-		if err := s.stream(context.Background()); err != nil {
-			t.Fatalf("stream returned error: %v", err)
-		}
-	}
-	s.client.CloseIdleConnections()
-
-	if !waitForGoroutines(baseline) {
-		t.Fatalf("goroutines did not return to baseline %d after %d reconnects, got %d",
-			baseline, reconnects, runtime.NumGoroutine())
-	}
-}
-
-// updateHandler wants UPDATE messages and does nothing with them.
-type updateHandler struct{}
-
-func (updateHandler) WantsMessage(messageType string) bool { return messageType == "UPDATE" }
-func (updateHandler) HandleMessage(*ristypes.RISMessage)   {}
-func (updateHandler) QueueCapacity() int                   { return 10 }
-
-// TestStopBeforeMessageReachesQueues stops the streamer after the read loop
-// has checked for cancellation but before it hands the message to the handler
-// queues. That is the gap a stop from another goroutine can land in, and it
-// used to end in "send on closed channel". The raw handler runs in that gap on
-// the read loop itself, so calling Stop from it hits the gap every time.
-func TestStopBeforeMessageReachesQueues(t *testing.T) {
-	const line = `{"type":"ris_message","data":{"type":"UPDATE","peer":"192.0.2.1",` +
-		`"peer_asn":"64496","timestamp":1700000000}}` + "\n"
-	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) {
-		_, _ = io.WriteString(w, line)
-	}))
-	defer srv.Close()
-
-	s := New(logger.New(), metrics.New())
-	s.url = srv.URL
-	s.RegisterHandler(updateHandler{})
-	s.RegisterRawHandler(func(string) { s.Stop() })
-
-	// Start would run the stream in the background, where the test cannot
-	// wait for it. Setting cancel as Start does lets Stop cancel the stream
-	// run here instead.
-	ctx, cancel := context.WithCancel(context.Background())
-	s.cancel = cancel
-
-	if err := s.stream(ctx); !errors.Is(err, context.Canceled) {
-		t.Fatalf("stream returned %v, want %v", err, context.Canceled)
-	}
-
-	// A second Stop must not close the queues again.
-	s.Stop()
-}
-
-// settledGoroutineCount lets transient goroutines finish, then reports the
-// current count.
-func settledGoroutineCount() int {
-	prev := runtime.NumGoroutine()
-	for range 20 {
-		time.Sleep(10 * time.Millisecond)
-		cur := runtime.NumGoroutine()
-		if cur == prev {
-			return cur
-		}
-		prev = cur
-	}
-
-	return prev
-}
-
-// waitForGoroutines waits until the goroutine count drops to target or below.
-func waitForGoroutines(target int) bool {
-	for range 100 {
-		if runtime.NumGoroutine() <= target {
-			return true
-		}
-		time.Sleep(10 * time.Millisecond)
-	}
-
-	return false
-}

internal/templates/as_detail.html

@@ -4,7 +4,7 @@
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title>AS{{.ASN.ASN}} - {{.ASN.Handle}} - RouteWatch</title>
-    <style>
+    <style nonce="{{.CSPNonce}}">
         * {
             box-sizing: border-box;
         }
@@ -126,6 +126,26 @@
             margin: 0;
             color: #2c3e50;
         }
+        /* A click on a prefix list's heading folds or unfolds the list. Not
+           every browser shows its own marker on a flex element, so the page
+           hides it and draws a triangle before the heading. The triangle
+           points down while the list is open. */
+        summary.prefix-header {
+            cursor: pointer;
+            list-style: none;
+        }
+        summary.prefix-header::-webkit-details-marker {
+            display: none;
+        }
+        summary.prefix-header h2::before {
+            content: "\25B8";
+            display: inline-block;
+            margin-right: 10px;
+            transition: transform 0.2s;
+        }
+        details[open] > summary.prefix-header h2::before {
+            transform: rotate(90deg);
+        }
         .prefix-count {
             background: #e74c3c;
             color: white;
@@ -233,18 +253,20 @@
                 <div class="info-label">Peer ASNs</div>
                 <div class="info-value">{{.PeerCount}}</div>
             </div>
+            {{if not .ASN.FirstSeen.IsZero}}
             <div class="info-card">
                 <div class="info-label">First Seen</div>
                 <div class="info-value">{{.ASN.FirstSeen.Format "2006-01-02"}}</div>
             </div>
+            {{end}}
         </div>
 
         {{if .IPv4Prefixes}}
-        <div class="prefix-section">
+        <details class="prefix-section" open>
-            <div class="prefix-header">
+            <summary class="prefix-header">
                 <h2>IPv4 Prefixes</h2>
                 <span class="prefix-count">{{.IPv4Count}}</span>
-            </div>
+            </summary>
             <table class="prefix-table">
                 <thead>
                     <tr>
@@ -265,15 +287,15 @@
                     {{end}}
                 </tbody>
             </table>
-        </div>
+        </details>
         {{end}}
 
         {{if .IPv6Prefixes}}
-        <div class="prefix-section">
+        <details class="prefix-section" open>
-            <div class="prefix-header">
+            <summary class="prefix-header">
                 <h2>IPv6 Prefixes</h2>
                 <span class="prefix-count">{{.IPv6Count}}</span>
-            </div>
+            </summary>
             <table class="prefix-table">
                 <thead>
                     <tr>
@@ -294,7 +316,7 @@
                     {{end}}
                 </tbody>
             </table>
-        </div>
+        </details>
         {{end}}
 
         {{if eq .TotalCount 0}}

internal/templates/index.html

@@ -4,7 +4,7 @@
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title>RouteWatch - BGP Route Monitor</title>
-    <style>
+    <style nonce="{{.CSPNonce}}">
         * {
             box-sizing: border-box;
         }
@@ -107,9 +107,15 @@
         .stat-card.connected .stat-value {
             color: #27ae60;
         }
-        .stat-card.disconnected .stat-value {
+        .stat-card.disconnected .stat-value,
+        .stale {
             color: #e74c3c;
         }
+        /* The card has no room beside "Connected", so the spinner and "stale" go under it */
+        #freshness {
+            display: block;
+            font-size: 20px;
+        }
 
         /* Search section */
         .search-section {
@@ -167,6 +173,45 @@
             color: #95a5a6;
             margin-top: 8px;
         }
+        #asname-results p {
+            color: #7f8c8d;
+            margin-top: 12px;
+        }
+        #asname-results .search-message {
+            font-size: 13px;
+        }
+        #asname-results .error {
+            color: #c00;
+        }
+        .asname-list {
+            list-style: none;
+            margin: 12px 0 0 0;
+            padding: 0;
+            max-height: 400px;
+            overflow-y: auto;
+        }
+        .asname-list li {
+            border-bottom: 1px solid #eee;
+        }
+        .asname-list a {
+            display: block;
+            padding: 8px 4px;
+            color: #2c3e50;
+            text-decoration: none;
+            font-size: 14px;
+        }
+        .asname-list a:hover {
+            background: #f5f5f5;
+        }
+        .asname-list .asn {
+            font-weight: 600;
+            color: #3498db;
+        }
+        .asname-list .description {
+            display: block;
+            color: #7f8c8d;
+            font-size: 12px;
+        }
 
         /* IP Lookup result */
         .ip-result {
@@ -194,7 +239,8 @@
             cursor: pointer;
             font-size: 12px;
         }
-        .ip-result pre {
+        .ip-result pre,
+        #ip-api pre {
             background: #2c3e50;
             color: #ecf0f1;
             padding: 16px;
@@ -206,18 +252,30 @@
             max-height: 400px;
             overflow-y: auto;
         }
-        .ip-result .error {
+        .ip-result .error,
+        #ip-api .error {
             background: #fee;
             color: #c00;
             padding: 12px;
             border-radius: 6px;
             font-size: 14px;
         }
-        .ip-result .loading {
+        .ip-result .loading,
+        #ip-api .loading {
             color: #7f8c8d;
             font-style: italic;
         }
 
+        /* IP information API card: the command, then its answer */
+        #ip-api p {
+            margin: 0;
+            color: #7f8c8d;
+            font-size: 14px;
+        }
+        #ip-api pre {
+            margin-top: 12px;
+        }
+
         /* Footer */
         .footer {
             margin-top: 40px;
@@ -258,7 +316,10 @@
     <main class="main-content">
         <div class="stats-grid">
             <div class="stat-card" id="status-card">
-                <div class="stat-value" id="stat-status">-</div>
+                <div class="stat-value">
+                    <span id="stat-status">-</span>
+                    <span id="freshness"></span>
+                </div>
                 <div class="stat-label">Status</div>
             </div>
             <div class="stat-card">
@@ -299,7 +360,7 @@
                     <input type="text" id="asname-input" placeholder="e.g., Google, Cloudflare" autocomplete="off">
                     <button type="submit">Search</button>
                 </form>
-                <p class="search-hint">Search for autonomous systems by organization name</p>
+                <p class="search-hint">Search for autonomous systems by organization name or handle</p>
                 <div id="asname-results"></div>
             </div>
 
@@ -319,6 +380,13 @@
                 </div>
             </div>
         </div>
+
+        <div class="search-card" id="ip-api">
+            <h2>IP Information API</h2>
+            <p>The API answers what RouteWatch knows about an IP address, such as its AS number and country. <code>/ip/</code> with no address answers for the caller's own address.</p>
+            <pre id="ip-api-command" data-client-ip="{{.ClientIP}}"></pre>
+            <pre id="ip-api-answer" class="loading">Loading...</pre>
+        </div>
     </main>
 
     <footer class="footer">
@@ -329,7 +397,45 @@
         <span><a href="{{appGitCommitURL}}">{{appGitRevision}}</a></span>
     </footer>
 
-    <script>
+    <script nonce="{{.CSPNonce}}">
+        // The stats are fetched every REFRESH_PERIOD_MS. Next to the status, a
+        // spinner turns one frame each time an answer counts more messages
+        // than the one before. After STALE_AFTER_REFRESHES refreshes without
+        // such an answer, the spinner gives way to "stale" and the time since.
+        const REFRESH_PERIOD_MS = 5000;
+        const STALE_AFTER_REFRESHES = 3;
+        const MS_PER_SECOND = 1000;
+        const SPINNER_FRAMES = ['⠋', '⠙', '⠹', '⠸', '⠼', '⠴', '⠦', '⠧', '⠇', '⠏'];
+        const SPINNER_COLOURS = ['#e74c3c', '#e67e22', '#f1c40f', '#27ae60', '#3498db', '#9b59b6'];
+
+        let spinnerFrame = 0;
+        let lastMessageCount = null;
+        let lastFreshAt = Date.now();
+
+        // Turn the spinner when the message count has risen since the last answer.
+        function noteMessageCount(count) {
+            if (lastMessageCount !== null && count > lastMessageCount) {
+                spinnerFrame = (spinnerFrame + 1) % SPINNER_FRAMES.length;
+                lastFreshAt = Date.now();
+            }
+            lastMessageCount = count;
+            showFreshness();
+        }
+
+        function showFreshness() {
+            const freshness = document.getElementById('freshness');
+            const sinceFreshMs = Date.now() - lastFreshAt;
+            const stale = sinceFreshMs >= STALE_AFTER_REFRESHES * REFRESH_PERIOD_MS;
+            freshness.classList.toggle('stale', stale);
+            if (stale) {
+                freshness.textContent = 'stale for ' + Math.floor(sinceFreshMs / MS_PER_SECOND) + 's';
+                freshness.style.color = '';
+                return;
+            }
+            freshness.textContent = SPINNER_FRAMES[spinnerFrame];
+            freshness.style.color = SPINNER_COLOURS[spinnerFrame % SPINNER_COLOURS.length];
+        }
+
         function formatNumber(num) {
             if (num >= 1000000) {
                 return (num / 1000000).toFixed(1) + 'M';
@@ -351,6 +457,7 @@
                     const statusEl = document.getElementById('stat-status');
                     statusEl.textContent = data.connected ? 'Connected' : 'Disconnected';
                     statusCard.className = 'stat-card ' + (data.connected ? 'connected' : 'disconnected');
+                    noteMessageCount(data.total_messages);
 
                     document.getElementById('stat-routes').textContent = formatNumber(data.live_routes);
                     document.getElementById('stat-asns').textContent = formatNumber(data.asns);
@@ -361,7 +468,7 @@
                     }
 
                     const totalUpdates = data.ipv4_updates_per_sec + data.ipv6_updates_per_sec;
-                    document.getElementById('stat-updates').textContent = totalUpdates.toFixed(1);
+                    document.getElementById('stat-updates').textContent = Math.round(totalUpdates).toLocaleString();
                 })
                 .catch(() => {
                     document.getElementById('stat-status').textContent = 'Error';
@@ -380,24 +487,82 @@
             }
         });
 
-        // AS name search
+        // AS name search. The handles and descriptions come from outside data,
+        // so the results are built with textContent, never innerHTML.
+        const asNameResults = document.getElementById('asname-results');
+
+        // Replace the search results with one line of text.
+        function showASNameMessage(text, className) {
+            const line = document.createElement('p');
+            line.className = className;
+            line.textContent = text;
+            asNameResults.replaceChildren(line);
+        }
+
+        // List the matches of a search answer, each as a link to its AS page,
+        // and say how many more matched than the answer holds.
+        function showASNameMatches(data) {
+            if (data.total === 0) {
+                showASNameMessage('No AS matches "' + data.query + '".', 'search-message');
+                return;
+            }
+
+            const list = document.createElement('ul');
+            list.className = 'asname-list';
+            for (const match of data.results) {
+                const asn = document.createElement('span');
+                asn.className = 'asn';
+                asn.textContent = 'AS' + match.asn;
+
+                const description = document.createElement('span');
+                description.className = 'description';
+                description.textContent = match.description;
+
+                const link = document.createElement('a');
+                link.href = '/as/' + match.asn;
+                link.append(asn, ' ' + match.handle, description);
+
+                const item = document.createElement('li');
+                item.append(link);
+                list.append(item);
+            }
+            asNameResults.replaceChildren(list);
+
+            const more = data.total - data.results.length;
+            if (more > 0) {
+                const line = document.createElement('p');
+                line.className = 'search-message';
+                line.textContent = more.toLocaleString() + ' more matched. Narrow the search to see them.';
+                asNameResults.append(line);
+            }
+        }
+
         document.getElementById('asname-form').addEventListener('submit', function(e) {
             e.preventDefault();
             const query = document.getElementById('asname-input').value.trim();
             if (!query) return;
 
-            const resultsDiv = document.getElementById('asname-results');
+            // A number, or AS and a number, opens that AS's page
-            resultsDiv.innerHTML = '<p class="loading" style="color: #7f8c8d; margin-top: 12px;">Searching...</p>';
+            const asn = query.replace(/^AS/i, '');
-
+            if (/^\d+$/.test(asn)) {
-            // Use a simple client-side search against the asinfo data
+                window.location.href = '/as/' + asn;
-            // For now, redirect to AS page if it looks like an ASN
-            if (/^\d+$/.test(query)) {
-                window.location.href = '/as/' + query;
                 return;
             }
 
-            // Show a message that server-side search is coming
+            showASNameMessage('Searching...', 'loading');
-            resultsDiv.innerHTML = '<p style="color: #7f8c8d; margin-top: 12px; font-size: 13px;">AS name search coming soon. For now, try an AS number.</p>';
+
+            fetch('/api/v1/search/as?q=' + encodeURIComponent(query))
+                .then(response => response.json())
+                .then(response => {
+                    if (response.status === 'error') {
+                        showASNameMessage('Error: ' + response.error.msg, 'error');
+                    } else {
+                        showASNameMatches(response.data);
+                    }
+                })
+                .catch(error => {
+                    showASNameMessage('Error: ' + error.message, 'error');
+                });
         });
 
         // IP lookup
@@ -439,9 +604,33 @@
             document.getElementById('ip-input').value = '';
         });
 
-        // Initial load and refresh stats every 5 seconds
+        // IP information API card: show the curl command that asks /ip/ about
+        // the visitor's own address, then make that same request and show the
+        // JSON it answers, an error answer included. The answer holds outside
+        // data, so it is set with textContent, never innerHTML.
+        const ipAPICommand = document.getElementById('ip-api-command');
+        const ipAPIAnswer = document.getElementById('ip-api-answer');
+        const ipAPIURL = window.location.origin + '/ip/' + ipAPICommand.dataset.clientIp;
+
+        ipAPICommand.textContent = 'curl ' + ipAPIURL;
+
+        fetch(ipAPIURL)
+            .then(response => response.json())
+            .then(answer => {
+                ipAPIAnswer.className = '';
+                ipAPIAnswer.textContent = JSON.stringify(answer, null, 2);
+            })
+            .catch(error => {
+                ipAPIAnswer.className = 'error';
+                ipAPIAnswer.textContent = 'Error: ' + error.message;
+            });
+
+        // Initial load and refresh stats every REFRESH_PERIOD_MS. The freshness
+        // check runs every second on its own, so a request that never returns
+        // still turns the spinner into "stale".
         updateStats();
-        setInterval(updateStats, 5000);
+        setInterval(updateStats, REFRESH_PERIOD_MS);
+        setInterval(showFreshness, MS_PER_SECOND);
     </script>
 </body>
 </html>

internal/templates/prefix_detail.html

@@ -4,7 +4,7 @@
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title>{{.Prefix}} - RouteWatch</title>
-    <style>
+    <style nonce="{{.CSPNonce}}">
         * {
             box-sizing: border-box;
         }
@@ -221,6 +221,10 @@
             border-radius: 4px;
             border: 1px solid #e0e0e0;
         }
+        .origin-item .peer-count {
+            color: #7f8c8d;
+            margin-left: 10px;
+        }
         .empty-state {
             text-align: center;
             padding: 40px;
@@ -281,9 +285,14 @@
             <div class="origin-list">
                 {{range .Origins}}
                 <div class="origin-item">
+                    {{if .ASN}}
                     <a href="/as/{{.ASN}}" class="as-link">AS{{.ASN}}</a>
                     {{if .Handle}} ({{.Handle}}){{end}}
-                    <span style="color: #7f8c8d; margin-left: 10px;">{{.PeerCount}} peer{{if ne .PeerCount 1}}s{{end}}</span>
+                    {{if .Description}} {{.Description}}{{end}}
+                    {{else}}
+                    No known origin
+                    {{end}}
+                    <span class="peer-count">{{.PeerCount}} peer{{if ne .PeerCount 1}}s{{end}}</span>
                 </div>
                 {{end}}
             </div>
@@ -311,7 +320,11 @@
                     {{range .Routes}}
                     <tr>
                         <td>
+                            {{if .OriginASN}}
                             <a href="/as/{{.OriginASN}}" class="as-link">AS{{.OriginASN}}</a>
+                            {{else}}
+                            No known origin
+                            {{end}}
                         </td>
                         <td class="peer-ip">{{.PeerIP}}</td>
                         <td class="as-path">{{range $i, $as := .ASPathWithHandle}}{{if $i}} → {{end}}<a href="/as/{{$as.ASN}}" class="as-link">{{if $as.Handle}}{{$as.Handle}}{{else}}AS{{$as.ASN}}{{end}}</a>{{end}}</td>

internal/templates/prefix_length.html

@@ -4,7 +4,7 @@
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title>Prefixes with /{{ .MaskLength }} - RouteWatch</title>
-    <style>
+    <style nonce="{{.CSPNonce}}">
         * {
             box-sizing: border-box;
         }
@@ -172,9 +172,13 @@
                 <td><a href="{{ .Prefix | prefixURL }}" class="prefix-link">{{ .Prefix }}</a></td>
                 <td class="age">{{ .Age }}</td>
                 <td>
+                    {{ if .OriginASN }}
                     <a href="/as/{{ .OriginASN }}" class="as-link">
                         AS{{ .OriginASN }}{{ if .OriginASDescription }} ({{ .OriginASDescription }}){{ end }}
                     </a>
+                    {{ else }}
+                    No known origin
+                    {{ end }}
                 </td>
             </tr>
             {{ end }}

internal/templates/status.html

@@ -4,7 +4,7 @@
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title>RouteWatch Status</title>
-    <style>
+    <style nonce="{{.CSPNonce}}">
         * {
             box-sizing: border-box;
         }
@@ -126,10 +126,12 @@
         .connected {
             color: #22c55e;
         }
-        .disconnected {
+        .disconnected,
+        .stale {
             color: #ef4444;
         }
         .error {
+            display: none;
             background: #fee;
             color: #c00;
             padding: 10px;
@@ -173,13 +175,16 @@
     </nav>
 
     <main class="main-content">
-    <div id="error" class="error" style="display: none;"></div>
+    <div id="error" class="error"></div>
     <div class="status-grid">
         <div class="status-card">
             <h2>RouteWatch</h2>
             <div class="metric">
                 <span class="metric-label">Status</span>
+                <span>
                     <span class="metric-value" id="connected">-</span>
+                    <span class="metric-value" id="freshness"></span>
+                </span>
             </div>
             <div class="metric">
                 <span class="metric-label">Uptime</span>
@@ -230,8 +235,8 @@
                 <span class="metric-value" id="withdrawals">-</span>
             </div>
             <div class="metric">
-                <span class="metric-label">Route Churn/sec</span>
+                <span class="metric-label">Route Churn/min</span>
-                <span class="metric-value" id="route_churn_per_sec">-</span>
+                <span class="metric-value" id="route_churn_per_min">-</span>
             </div>
             <div class="metric">
                 <span class="metric-label">Total Data</span>
@@ -384,7 +389,45 @@
         <!-- Handler stats will be dynamically added here -->
     </div>
 
-    <script>
+    <script nonce="{{.CSPNonce}}">
+        // The stats are fetched every REFRESH_PERIOD_MS. Next to the status, a
+        // spinner turns one frame each time an answer counts more messages
+        // than the one before. After STALE_AFTER_REFRESHES refreshes without
+        // such an answer, the spinner gives way to "stale" and the time since.
+        const REFRESH_PERIOD_MS = 2000;
+        const STALE_AFTER_REFRESHES = 3;
+        const MS_PER_SECOND = 1000;
+        const SPINNER_FRAMES = ['⠋', '⠙', '⠹', '⠸', '⠼', '⠴', '⠦', '⠧', '⠇', '⠏'];
+        const SPINNER_COLOURS = ['#e74c3c', '#e67e22', '#f1c40f', '#27ae60', '#3498db', '#9b59b6'];
+
+        let spinnerFrame = 0;
+        let lastMessageCount = null;
+        let lastFreshAt = Date.now();
+
+        // Turn the spinner when the message count has risen since the last answer.
+        function noteMessageCount(count) {
+            if (lastMessageCount !== null && count > lastMessageCount) {
+                spinnerFrame = (spinnerFrame + 1) % SPINNER_FRAMES.length;
+                lastFreshAt = Date.now();
+            }
+            lastMessageCount = count;
+            showFreshness();
+        }
+
+        function showFreshness() {
+            const freshness = document.getElementById('freshness');
+            const sinceFreshMs = Date.now() - lastFreshAt;
+            const stale = sinceFreshMs >= STALE_AFTER_REFRESHES * REFRESH_PERIOD_MS;
+            freshness.classList.toggle('stale', stale);
+            if (stale) {
+                freshness.textContent = 'stale for ' + Math.floor(sinceFreshMs / MS_PER_SECOND) + 's';
+                freshness.style.color = '';
+                return;
+            }
+            freshness.textContent = SPINNER_FRAMES[spinnerFrame];
+            freshness.style.color = SPINNER_COLOURS[spinnerFrame % SPINNER_COLOURS.length];
+        }
+
         function formatBytes(bytes) {
             if (bytes === 0) return '0 B';
             const k = 1024;
@@ -510,7 +553,7 @@
             document.getElementById('messages_per_sec').textContent = '-';
             document.getElementById('announcements').textContent = '-';
             document.getElementById('withdrawals').textContent = '-';
-            document.getElementById('route_churn_per_sec').textContent = '-';
+            document.getElementById('route_churn_per_min').textContent = '-';
             document.getElementById('total_wire_bytes').textContent = '-';
             document.getElementById('wire_mbits_per_sec').textContent = '-';
             document.getElementById('gc_num').textContent = '-';
@@ -568,6 +611,7 @@
                     const connectedEl = document.getElementById('connected');
                     connectedEl.textContent = data.connected ? 'Connected' : 'Disconnected';
                     connectedEl.className = 'metric-value ' + (data.connected ? 'connected' : 'disconnected');
+                    noteMessageCount(data.total_messages);
 
                     // Update all metrics
                     document.getElementById('uptime').textContent = data.uptime;
@@ -577,7 +621,7 @@
                     document.getElementById('connection_duration').textContent = data.connection_duration;
                     document.getElementById('reconnect_count').textContent = formatNumber(data.reconnect_count);
                     document.getElementById('total_messages').textContent = formatNumber(data.total_messages);
-                    document.getElementById('messages_per_sec').textContent = data.messages_per_sec.toFixed(1);
+                    document.getElementById('messages_per_sec').textContent = formatNumber(Math.round(data.messages_per_sec));
                     document.getElementById('total_wire_bytes').textContent = formatBytes(data.total_wire_bytes);
                     document.getElementById('wire_mbits_per_sec').textContent = data.wire_mbits_per_sec.toFixed(2) + ' Mbps';
                     document.getElementById('asns').textContent = formatNumber(data.asns);
@@ -590,8 +634,8 @@
                     document.getElementById('live_routes').textContent = formatNumber(data.live_routes);
                     document.getElementById('ipv4_routes').textContent = formatNumber(data.ipv4_routes);
                     document.getElementById('ipv6_routes').textContent = formatNumber(data.ipv6_routes);
-                    document.getElementById('ipv4_updates_per_sec').textContent = data.ipv4_updates_per_sec.toFixed(1);
+                    document.getElementById('ipv4_updates_per_sec').textContent = formatNumber(Math.round(data.ipv4_updates_per_sec));
-                    document.getElementById('ipv6_updates_per_sec').textContent = data.ipv6_updates_per_sec.toFixed(1);
+                    document.getElementById('ipv6_updates_per_sec').textContent = formatNumber(Math.round(data.ipv6_updates_per_sec));
                     document.getElementById('oldest_route').textContent = formatRelativeTime(data.oldest_route);
                     document.getElementById('newest_route').textContent = formatRelativeTime(data.newest_route);
 
@@ -600,7 +644,7 @@
                         document.getElementById('bgp_peer_count').textContent = formatNumber(data.stream.bgp_peer_count);
                         document.getElementById('announcements').textContent = formatNumber(data.stream.announcements);
                         document.getElementById('withdrawals').textContent = formatNumber(data.stream.withdrawals);
-                        document.getElementById('route_churn_per_sec').textContent = data.stream.route_churn_per_sec.toFixed(1);
+                        document.getElementById('route_churn_per_min').textContent = formatNumber(data.stream.route_churn_per_min);
                     }
 
                     // Update GC stats
@@ -642,9 +686,12 @@
                 });
         }
         
-        // Update immediately and then every 2 seconds
+        // Update immediately and then every REFRESH_PERIOD_MS. The freshness
+        // check runs every second on its own, so a request that never returns
+        // still turns the spinner into "stale".
         updateStatus();
-        setInterval(updateStatus, 2000);
+        setInterval(updateStatus, REFRESH_PERIOD_MS);
+        setInterval(showFreshness, MS_PER_SECOND);
     </script>
     </main>
 

internal/templates/templates.go

@@ -7,9 +7,9 @@ import (
 	"net/url"
 	"strings"
 	"sync"
-	"time"
 
 	"git.eeqj.de/sneak/routewatch/internal/version"
+	"github.com/dustin/go-humanize"
 )
 
 //go:embed index.html
@@ -48,44 +48,7 @@ var (
 	once sync.Once
 )
 
-const (
+const cidrPartCount = 2 // A CIDR has two parts: prefix and length
-	hoursPerDay   = 24
-	daysPerMonth  = 30
-	cidrPartCount = 2 // A CIDR has two parts: prefix and length
-)
-
-// timeSince returns a human-readable duration since the given time
-func timeSince(t time.Time) string {
-	duration := time.Since(t)
-	if duration < time.Minute {
-		return "just now"
-	}
-	if duration < time.Hour {
-		minutes := int(duration.Minutes())
-		if minutes == 1 {
-			return "1 minute ago"
-		}
-
-		return duration.Truncate(time.Minute).String() + " ago"
-	}
-	if duration < hoursPerDay*time.Hour {
-		hours := int(duration.Hours())
-		if hours == 1 {
-			return "1 hour ago"
-		}
-
-		return duration.Truncate(time.Hour).String() + " ago"
-	}
-	days := int(duration.Hours() / hoursPerDay)
-	if days == 1 {
-		return "1 day ago"
-	}
-	if days < daysPerMonth {
-		return duration.Truncate(hoursPerDay*time.Hour).String() + " ago"
-	}
-
-	return t.Format("2006-01-02")
-}
 
 // prefixURL generates a URL path for a prefix in CIDR notation.
 // Takes a prefix like "192.168.1.0/24" and returns "/prefix/192.168.1.0/24"
@@ -109,7 +72,7 @@ func initTemplates() {
 
 	// Create common template functions
 	funcs := template.FuncMap{
-		"timeSince":       timeSince,
+		"timeSince":       humanize.Time,
 		"urlEncode":       url.QueryEscape,
 		"prefixURL":       prefixURL,
 		"appName":         func() string { return version.Name },
@@ -118,7 +81,7 @@ func initTemplates() {
 		"appLicense":      func() string { return version.License },
 		"appRepoURL":      func() string { return version.RepoURL },
 		"appGitRevision":  func() string { return version.GitRevisionShort },
-		"appGitCommitURL": func() string { return version.CommitURL() },
+		"appGitCommitURL": version.CommitURL,
 	}
 
 	// Parse index template
@@ -134,19 +97,22 @@ func initTemplates() {
 	}
 
 	// Parse AS detail template
-	defaultTemplates.ASDetail, err = template.New("asDetail").Funcs(funcs).Parse(asDetailHTML)
+	defaultTemplates.ASDetail, err = template.New("asDetail").
+		Funcs(funcs).Parse(asDetailHTML)
 	if err != nil {
 		panic("failed to parse AS detail template: " + err.Error())
 	}
 
 	// Parse prefix detail template
-	defaultTemplates.PrefixDetail, err = template.New("prefixDetail").Funcs(funcs).Parse(prefixDetailHTML)
+	defaultTemplates.PrefixDetail, err = template.New("prefixDetail").
+		Funcs(funcs).Parse(prefixDetailHTML)
 	if err != nil {
 		panic("failed to parse prefix detail template: " + err.Error())
 	}
 
 	// Parse prefix length template
-	defaultTemplates.PrefixLength, err = template.New("prefixLength").Funcs(funcs).Parse(prefixLengthHTML)
+	defaultTemplates.PrefixLength, err = template.New("prefixLength").
+		Funcs(funcs).Parse(prefixLengthHTML)
 	if err != nil {
 		panic("failed to parse prefix length template: " + err.Error())
 	}

internal/templates/templates_test.go

@@ -0,0 +1,181 @@
+package templates_test
+
+import (
+	"bytes"
+	"html/template"
+	"strings"
+	"testing"
+
+	"git.eeqj.de/sneak/routewatch/internal/database"
+	"git.eeqj.de/sneak/routewatch/internal/templates"
+)
+
+// renderPage renders a page with the data every page needs, and the visitor's
+// address, which the home page also needs.
+func renderPage(t *testing.T, tmpl *template.Template) string {
+	t.Helper()
+
+	var page bytes.Buffer
+
+	err := tmpl.Execute(&page, struct{ CSPNonce, ClientIP string }{})
+	if err != nil {
+		t.Fatalf("rendering the page: %v", err)
+	}
+
+	return page.String()
+}
+
+// TestFooterLicense renders the home and status pages and expects each footer
+// to name the licence in LICENSE, MIT, and not WTFPL.
+func TestFooterLicense(t *testing.T) {
+	t.Parallel()
+
+	const (
+		license      = "MIT"
+		wrongLicense = "WTFPL"
+	)
+
+	pages := map[string]*template.Template{
+		"index":  templates.IndexTemplate(),
+		"status": templates.StatusTemplate(),
+	}
+
+	for name, tmpl := range pages {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			_, footer, found := strings.Cut(renderPage(t, tmpl), "<footer")
+			if !found {
+				t.Fatal("the page has no footer")
+			}
+
+			footer, _, _ = strings.Cut(footer, "</footer>")
+
+			if !strings.Contains(footer, license) {
+				t.Errorf("footer does not name %s: %q", license, footer)
+			}
+
+			if strings.Contains(footer, wrongLicense) {
+				t.Errorf("footer names %s: %q", wrongLicense, footer)
+			}
+		})
+	}
+}
+
+// TestFreshnessIndicator renders the home and status pages and expects each to
+// have, right after the status element, the element where the page's script
+// shows its spinner or "stale".
+func TestFreshnessIndicator(t *testing.T) {
+	t.Parallel()
+
+	pages := map[string]struct {
+		tmpl      *template.Template
+		status    string
+		indicator string
+	}{
+		"index": {
+			tmpl:      templates.IndexTemplate(),
+			status:    `<span id="stat-status">-</span>`,
+			indicator: `<span id="freshness"></span>`,
+		},
+		"status": {
+			tmpl:      templates.StatusTemplate(),
+			status:    `<span class="metric-value" id="connected">-</span>`,
+			indicator: `<span class="metric-value" id="freshness"></span>`,
+		},
+	}
+
+	for name, page := range pages {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			_, afterStatus, found := strings.Cut(renderPage(t, page.tmpl), page.status)
+			if !found {
+				t.Fatalf("the page has no status element %s", page.status)
+			}
+
+			if !strings.HasPrefix(strings.TrimSpace(afterStatus), page.indicator) {
+				t.Errorf("the status element is not followed by %s", page.indicator)
+			}
+		})
+	}
+}
+
+// TestASDetailPrefixListsFold renders the AS page with one IPv4 and one IPv6
+// prefix. It expects each prefix list inside its own <details> element that
+// starts open and whose <summary> holds the list's heading, so that a click on
+// the heading folds the list.
+func TestASDetailPrefixListsFold(t *testing.T) {
+	t.Parallel()
+
+	const (
+		asn      = 64500
+		prefixV4 = "192.0.2.0/24"
+		prefixV6 = "2001:db8::/32"
+	)
+
+	ipv4 := []database.LiveRoute{{Prefix: prefixV4}}
+	ipv6 := []database.LiveRoute{{Prefix: prefixV6}}
+
+	// The fields the AS page handler gives the template.
+	data := struct {
+		ASN          *database.ASN
+		IPv4Prefixes []database.LiveRoute
+		IPv6Prefixes []database.LiveRoute
+		TotalCount   int
+		IPv4Count    int
+		IPv6Count    int
+		Peers        []database.ASPeer
+		PeerCount    int
+		CSPNonce     string
+	}{
+		ASN:          &database.ASN{ASN: asn},
+		IPv4Prefixes: ipv4,
+		IPv6Prefixes: ipv6,
+		TotalCount:   len(ipv4) + len(ipv6),
+		IPv4Count:    len(ipv4),
+		IPv6Count:    len(ipv6),
+	}
+
+	var rendered bytes.Buffer
+
+	err := templates.ASDetailTemplate().Execute(&rendered, data)
+	if err != nil {
+		t.Fatalf("rendering the page: %v", err)
+	}
+
+	page := rendered.String()
+
+	lists := []struct{ heading, prefix string }{
+		{"IPv4 Prefixes", prefixV4},
+		{"IPv6 Prefixes", prefixV6},
+	}
+
+	for _, list := range lists {
+		before, _, found := strings.Cut(page, list.prefix)
+		if !found {
+			t.Fatalf("the page does not list %s", list.prefix)
+		}
+
+		// The list's section starts at the last <details before its prefix.
+		start := strings.LastIndex(before, "<details")
+		if start < 0 || strings.Contains(before[start:], "</details>") {
+			t.Errorf("%s: the list is not inside a <details> element", list.heading)
+
+			continue
+		}
+
+		tag, section, _ := strings.Cut(before[start:], ">")
+		if !strings.Contains(tag, " open") {
+			t.Errorf("%s: the <details> element does not start open: %s>",
+				list.heading, tag)
+		}
+
+		_, summary, _ := strings.Cut(section, "<summary")
+		summary, _, _ = strings.Cut(summary, "</summary>")
+
+		if !strings.Contains(summary, list.heading) {
+			t.Errorf("%s: the <summary> does not hold the heading", list.heading)
+		}
+	}
+}

internal/version/version.go

@@ -3,7 +3,7 @@ package version
 
 // Build-time variables set via ldflags
 //
-//nolint:gochecknoglobals // These must be variables to allow ldflags injection at build time
+//nolint:gochecknoglobals // variables so that ldflags can set them at build time
 var (
 	// GitRevision is the git commit hash
 	GitRevision = "unknown"
@@ -19,8 +19,8 @@ const (
 	Author = "@sneak"
 	// AuthorURL is the author's website
 	AuthorURL = "https://sneak.berlin"
-	// License is the program license
+	// License is the program license, as LICENSE grants it
-	License = "WTFPL"
+	License = "MIT"
 	// RepoURL is the git repository URL
 	RepoURL = "https://git.eeqj.de/sneak/routewatch"
 )

internal/whois/whois.go

@@ -4,11 +4,18 @@ package whois
 import (
 	"bufio"
 	"context"
+	"errors"
 	"fmt"
+	"io"
+	"log/slog"
 	"net"
 	"regexp"
 	"strings"
+	"sync"
 	"time"
+	"unicode/utf8"
+
+	"golang.org/x/text/encoding/charmap"
 )
 
 // Timeout constants for WHOIS queries.
@@ -18,12 +25,24 @@ const (
 	writeTimeout = 5 * time.Second
 )
 
+// maxAnswerSize caps the bytes read of one WHOIS answer: a real one is a
+// few KiB, so one this large is from a server that does not stop sending.
+const maxAnswerSize = 1 << 20
+
 // Parsing constants.
 const (
-	keyValueParts       = 2 // Expected parts when splitting "key: value"
 	lacnicDateFormatLen = 8 // Length of YYYYMMDD date format
 )
 
+// The classes of the two objects that describe an AS in an answer from
+// RIPE, APNIC or AFRINIC, and the key of the aut-num object's line that
+// gives the ID of the AS's organisation object.
+const (
+	autNumClass       = "aut-num"
+	organisationClass = "organisation"
+	orgKey            = "org"
+)
+
 // WHOIS server addresses.
 const (
 	whoisServerIANA    = "whois.iana.org:43"
@@ -32,8 +51,32 @@ const (
 	whoisServerAPNIC   = "whois.apnic.net:43"
 	whoisServerLACNIC  = "whois.lacnic.net:43"
 	whoisServerAFRINIC = "whois.afrinic.net:43"
+
+	// NIC.br registers Brazil's ASes under LACNIC.
+	whoisServerNICbr = "whois.registro.br:43"
 )
 
+// nicbrHeader starts the copyright notice at the top of NIC.br's WHOIS
+// answers. LACNIC's answer for an AS that NIC.br registers is NIC.br's, and
+// carries the same notice:
+//
+//	% Copyright (c) Nic.br - Use of this data is governed by the Use and
+//	% Privacy Policy at https://registro.br/upp . Distribution,
+const nicbrHeader = "% Copyright (c) Nic.br"
+
+// nicbrLimitReached is in the line that marks NIC.br's reduced answer to a
+// client over its query limit. NIC.br publishes no limit. Its Use and
+// Privacy Policy, https://registro.br/upp, says only that it restricts
+// queries to stop the data being taken in bulk.
+const nicbrLimitReached = "Query rate limit exceeded"
+
+// NICbrSkipTime is how long the client stops asking NIC.br after NIC.br
+// answers that the client is over its query limit. NIC.br publishes
+// neither the limit nor how long it lasts. An hour outlasts a limit
+// counted per minute or per hour. The ASes looked up meanwhile get
+// NIC.br's fields when they are next refreshed.
+const NICbrSkipTime = time.Hour
+
 // RIR identifiers.
 const (
 	RIRARIN   = "ARIN"
@@ -43,6 +86,45 @@ const (
 	RIRAFRNIC = "AFRINIC"
 )
 
+// noPersonalFlag keeps person and role objects out of an answer. The
+// organisation object stays in it. RIPE counts the person and role objects
+// it returns against a daily limit per IP address. Role objects for abuse
+// contacts do not count. RIPE refuses an address for good that passes the
+// limit too often:
+// https://docs.db.ripe.net/Types-of-Queries/Access-Control-for-Queries/
+//
+// RIPE documents the flag here:
+// https://docs.db.ripe.net/Types-of-Queries/Referenced-Objects-in-Query-Response/
+// APNIC and AFRINIC list it in their answers to the query "help". AFRINIC
+// sends readers to that answer for its full list of flags:
+// https://afrinic.net/whois-getting-started
+const noPersonalFlag = "--no-personal"
+
+// The error lines with which RIPE's database software refuses a query:
+// access denied to the client's IP address, and that address's limit
+// reached. APNIC and AFRINIC run the same software. The lines are listed
+// under "Access Errors" on this page, whose URL is split over two lines:
+// https://docs.db.ripe.net/Appendices/
+// Appendix-C--RIPE-Database-Query-Server-Response-Codes-and-Messages/
+const (
+	accessDeniedError = "%ERROR:201:"
+	accessLimitError  = "%ERROR:202:"
+)
+
+// Errors LookupASN returns when a WHOIS answer has nothing to store.
+var (
+	// ErrNoReferral is returned when IANA refers the AS to no RIR, as it
+	// does for a reserved, private-use or unallocated AS number.
+	ErrNoReferral = errors.New("IANA refers the AS to no RIR")
+
+	errNoASBlock = errors.New("IANA answered with neither a referral " +
+		"nor an AS block")
+	errNoFields       = errors.New("RIR answered with no WHOIS fields")
+	errRefused        = errors.New("RIR refused the query")
+	errUnknownServer  = errors.New("IANA refers the AS to a server no RIR runs")
+	errAnswerTooLarge = errors.New("WHOIS answer reached the size cap")
+)
+
 // ASNInfo contains parsed WHOIS information for an ASN.
 type ASNInfo struct {
 	ASN         int
@@ -63,20 +145,46 @@ type ASNInfo struct {
 
 // Client performs WHOIS lookups for ASNs.
 type Client struct {
-	// Dialer for creating connections (can be overridden for testing)
+	// DialContext opens the connection to a WHOIS server, given as
-	dialer *net.Dialer
+	// host:port. NewClient sets a net.Dialer's; tests replace it to reach
+	// local servers instead.
+	DialContext func(ctx context.Context, network, address string) (net.Conn, error)
+
+	// Now returns the current time. NewClient sets time.Now; tests replace
+	// it to move the clock.
+	Now func() time.Time
+
+	logger *slog.Logger
+
+	// mu guards nicbrSkipUntil.
+	mu sync.Mutex
+
+	// nicbrSkipUntil is when the client next asks NIC.br, after NIC.br
+	// answered that the client is over its query limit.
+	nicbrSkipUntil time.Time
 }
 
-// NewClient creates a new WHOIS client.
+// NewClient creates a new WHOIS client that logs to logger.
-func NewClient() *Client {
+func NewClient(logger *slog.Logger) *Client {
+	dialer := &net.Dialer{Timeout: dialTimeout}
+
 	return &Client{
-		dialer: &net.Dialer{
+		DialContext: dialer.DialContext,
-			Timeout: dialTimeout,
+		Now:         time.Now,
-		},
+		logger:      logger,
 	}
 }
 
-// LookupASN queries WHOIS for the given ASN and returns parsed information.
+// LookupASN asks IANA which RIR holds the given ASN, asks that RIR, and
+// returns what the RIR's answer says. When LACNIC's answer is NIC.br's, it
+// returns what NIC.br's own server says instead, unless that query fails or
+// is skipped, as askNICbr says.
+//
+// It returns ErrNoReferral when IANA refers the AS to no RIR, and another
+// error when the query to IANA or to the RIR fails, when the RIR refuses
+// it, when IANA's answer has neither a referral nor the AS's block, when
+// IANA refers the AS to a server none of the five RIRs run, or when the
+// answer carries none of ASNInfo's fields.
 func (c *Client) LookupASN(ctx context.Context, asn int) (*ASNInfo, error) {
 	// Query IANA first to find the authoritative RIR
 	query := fmt.Sprintf("AS%d", asn)
@@ -87,72 +195,216 @@ func (c *Client) LookupASN(ctx context.Context, asn int) (*ASNInfo, error) {
 	}
 
 	// Determine RIR from IANA response
-	rir, whoisServer := c.parseIANAReferral(ianaResp)
+	rir, whoisServer, err := c.parseIANAReferral(ianaResp)
+	if err != nil {
+		return nil, fmt.Errorf("%s: %w", query, err)
+	}
+
 	if whoisServer == "" {
-		// No referral, try to parse what we have
+		// IANA names the block of an AS it refers to no RIR. An answer
-		return c.parseResponse(asn, rir, ianaResp), nil
+		// without that, such as an empty one, says nothing about the AS.
+		if !strings.Contains(ianaResp, "as-block:") {
+			return nil, fmt.Errorf("%s: %w", query, errNoASBlock)
+		}
+
+		return nil, fmt.Errorf("%s: %w", query, ErrNoReferral)
 	}
 
 	// Query the authoritative RIR
-	rirResp, err := c.query(ctx, whoisServer, query)
+	rirResp, err := c.queryRIR(ctx, whoisServer, rirQuery(rir, query))
 	if err != nil {
-		// Return partial data from IANA if RIR query fails
+		return nil, fmt.Errorf("RIR query failed: %w", err)
-		info := c.parseResponse(asn, rir, ianaResp)
+	}
-		info.RawResponse = ianaResp + "\n--- RIR query failed: " + err.Error() + " ---\n"
+
+	// The fields never come from IANA's answer, which describes the block
+	// the AS is in.
+	info := c.parseResponse(asn, rir, rirResp)
+
+	if rir == RIRLACNIC && strings.Contains(rirResp, nicbrHeader) {
+		info = c.askNICbr(ctx, query, info)
+	}
+
+	if info.isEmpty() {
+		return nil, fmt.Errorf("%s from %s: %w", query, whoisServer, errNoFields)
+	}
+
+	info.RawResponse = ianaResp + "\n" + info.RawResponse
 
 	return info, nil
 }
 
-	// Combine responses and parse
+// askNICbr asks NIC.br's server about an AS whose answer from LACNIC is
-	fullResponse := ianaResp + "\n" + rirResp
+// NIC.br's, and returns what NIC.br's answer says. LACNIC replaces each
-	info := c.parseResponse(asn, rir, fullResponse)
+// accented letter in such an answer with U+FFFD. NIC.br's own server keeps
-	info.RawResponse = fullResponse
+// them, in Latin-1.
+//
+// askNICbr returns lacnicInfo, what LACNIC's answer says, and logs a
+// warning when the query fails, when NIC.br sends the reduced answer for a
+// client over its limit, or when NIC.br's answer carries none of ASNInfo's
+// fields. After the reduced answer, it returns lacnicInfo without asking
+// NIC.br until NICbrSkipTime has passed.
+func (c *Client) askNICbr(
+	ctx context.Context, asQuery string, lacnicInfo *ASNInfo,
+) *ASNInfo {
+	asn := lacnicInfo.ASN
 
-	return info, nil
+	c.mu.Lock()
+	skipping := c.Now().Before(c.nicbrSkipUntil)
+	c.mu.Unlock()
+
+	if skipping {
+		return lacnicInfo
 	}
 
-// query performs a raw WHOIS query to the specified server.
+	answer, err := c.query(ctx, whoisServerNICbr, asQuery)
+	if err != nil {
+		c.logger.Warn("NIC.br query failed, keeping LACNIC's fields",
+			"asn", asn, "error", err)
+
+		return lacnicInfo
+	}
+
+	if strings.Contains(answer, nicbrLimitReached) {
+		c.mu.Lock()
+		c.nicbrSkipUntil = c.Now().Add(NICbrSkipTime)
+		c.mu.Unlock()
+
+		c.logger.Warn("NIC.br answered that the client is over its query "+
+			"limit, keeping LACNIC's fields and skipping NIC.br",
+			"asn", asn, "skip_for", NICbrSkipTime)
+
+		return lacnicInfo
+	}
+
+	info := c.parseResponse(asn, RIRLACNIC, answer)
+	if info.isEmpty() {
+		c.logger.Warn("NIC.br answered with no WHOIS fields, "+
+			"keeping LACNIC's fields", "asn", asn)
+
+		return lacnicInfo
+	}
+
+	return info
+}
+
+// isEmpty reports whether info holds none of the fields a WHOIS answer
+// fills, as when the answer is only comment and error lines.
+func (info *ASNInfo) isEmpty() bool {
+	return info.ASName == "" && info.OrgName == "" && info.OrgID == "" &&
+		info.Address == "" && info.CountryCode == "" &&
+		info.AbuseEmail == "" && info.AbusePhone == "" &&
+		info.TechEmail == "" && info.TechPhone == "" &&
+		info.RegDate == nil && info.LastMod == nil
+}
+
+// rirQuery returns the query sent to rir's server for asQuery, an AS
+// number such as "AS3333". RIPE, APNIC and AFRINIC are asked to keep person
+// and role objects out of the answer.
+func rirQuery(rir, asQuery string) string {
+	switch rir {
+	case RIRRIPE, RIRAPNIC, RIRAFRNIC:
+		return noPersonalFlag + " " + asQuery
+	default:
+		return asQuery
+	}
+}
+
+// queryRIR sends query to an RIR's server and returns the answer. An answer
+// holding an error line with which the server refuses the query is an
+// error, and the error holds that line.
+func (c *Client) queryRIR(
+	ctx context.Context, server, query string,
+) (string, error) {
+	answer, err := c.query(ctx, server, query)
+	if err != nil {
+		return "", err
+	}
+
+	for line := range strings.SplitSeq(answer, "\n") {
+		if strings.HasPrefix(line, accessDeniedError) ||
+			strings.HasPrefix(line, accessLimitError) {
+			return "", fmt.Errorf("%w: %s", errRefused, line)
+		}
+	}
+
+	return answer, nil
+}
+
+// query performs a raw WHOIS query to the specified server and returns the
+// answer in UTF-8. An answer that reaches maxAnswerSize is an error.
 func (c *Client) query(ctx context.Context, server, query string) (string, error) {
-	conn, err := c.dialer.DialContext(ctx, "tcp", server)
+	conn, err := c.DialContext(ctx, "tcp", server)
 	if err != nil {
 		return "", fmt.Errorf("dial %s: %w", server, err)
 	}
 	defer func() { _ = conn.Close() }()
 
 	// Set deadlines
-	if err := conn.SetWriteDeadline(time.Now().Add(writeTimeout)); err != nil {
+	err = conn.SetWriteDeadline(time.Now().Add(writeTimeout))
+	if err != nil {
 		return "", fmt.Errorf("set write deadline: %w", err)
 	}
 
 	// Send query
-	if _, err := fmt.Fprintf(conn, "%s\r\n", query); err != nil {
+	_, err = fmt.Fprintf(conn, "%s\r\n", query)
+	if err != nil {
 		return "", fmt.Errorf("write query: %w", err)
 	}
 
 	// Read response
-	if err := conn.SetReadDeadline(time.Now().Add(readTimeout)); err != nil {
+	err = conn.SetReadDeadline(time.Now().Add(readTimeout))
+	if err != nil {
 		return "", fmt.Errorf("set read deadline: %w", err)
 	}
 
 	var sb strings.Builder
-	scanner := bufio.NewScanner(conn)
+
+	// The reader ends at the cap; N, the bytes it has left, is 0 once the
+	// answer has reached it.
+	answer := &io.LimitedReader{R: conn, N: maxAnswerSize}
+
+	scanner := bufio.NewScanner(answer)
 	for scanner.Scan() {
 		sb.WriteString(scanner.Text())
 		sb.WriteString("\n")
 	}
 
-	if err := scanner.Err(); err != nil {
+	err = scanner.Err()
+	if err != nil {
 		return sb.String(), fmt.Errorf("read response: %w", err)
 	}
 
-	return sb.String(), nil
+	if answer.N == 0 {
+		return "", fmt.Errorf("read response: %w", errAnswerTooLarge)
+	}
+
+	return toUTF8(sb.String())
+}
+
+// toUTF8 returns a WHOIS answer in UTF-8. An answer that is valid UTF-8 is
+// returned as it is. Any other answer is decoded as Latin-1, which RIPE,
+// APNIC and NIC.br send. LACNIC, ARIN and AFRINIC send UTF-8. RIPE's
+// database working group lists the five RIRs here:
+// https://www.ripe.net/ripe/mail/archives/db-wg/2024-May/008075.html
+func toUTF8(answer string) (string, error) {
+	if utf8.ValidString(answer) {
+		return answer, nil
+	}
+
+	decoded, err := charmap.ISO8859_1.NewDecoder().String(answer)
+	if err != nil {
+		return "", fmt.Errorf("decode Latin-1 answer: %w", err)
+	}
+
+	return decoded, nil
 }
 
 // parseIANAReferral extracts the RIR and WHOIS server from an IANA response.
-func (c *Client) parseIANAReferral(response string) (rir, whoisServer string) {
+// The server is "" when the response has no referral. A referral to a server
-	lines := strings.Split(response, "\n")
+// that is none of the five RIRs' is an error, so a forged or broken answer
-
+// cannot send the client to any other host or port.
-	for _, line := range lines {
+func (c *Client) parseIANAReferral(response string) (string, string, error) {
+	for line := range strings.SplitSeq(response, "\n") {
 		line = strings.TrimSpace(line)
 
 		// Look for "refer:" line
@@ -162,47 +414,27 @@ func (c *Client) parseIANAReferral(response string) (rir, whoisServer string) {
 
 			switch {
 			case strings.Contains(server, "arin"):
-				return RIRARIN, whoisServerARIN
+				return RIRARIN, whoisServerARIN, nil
 			case strings.Contains(server, "ripe"):
-				return RIRRIPE, whoisServerRIPE
+				return RIRRIPE, whoisServerRIPE, nil
 			case strings.Contains(server, "apnic"):
-				return RIRAPNIC, whoisServerAPNIC
+				return RIRAPNIC, whoisServerAPNIC, nil
 			case strings.Contains(server, "lacnic"):
-				return RIRLACNIC, whoisServerLACNIC
+				return RIRLACNIC, whoisServerLACNIC, nil
 			case strings.Contains(server, "afrinic"):
-				return RIRAFRNIC, whoisServerAFRINIC
+				return RIRAFRNIC, whoisServerAFRINIC, nil
 			default:
-				// Unknown server, add port if missing
+				return "", "", fmt.Errorf("%w: %q", errUnknownServer, server)
-				if !strings.Contains(server, ":") {
-					server += ":43"
-				}
-
-				return "", server
-			}
-		}
-
-		// Also check organisation line for RIR hints
-		if strings.HasPrefix(strings.ToLower(line), "organisation:") {
-			org := strings.ToLower(line)
-			switch {
-			case strings.Contains(org, "arin"):
-				rir = RIRARIN
-			case strings.Contains(org, "ripe"):
-				rir = RIRRIPE
-			case strings.Contains(org, "apnic"):
-				rir = RIRAPNIC
-			case strings.Contains(org, "lacnic"):
-				rir = RIRLACNIC
-			case strings.Contains(org, "afrinic"):
-				rir = RIRAFRNIC
 			}
 		}
 	}
 
-	return rir, ""
+	return "", "", nil
 }
 
-// parseResponse extracts ASN information from a WHOIS response.
+// parseResponse extracts ASN information from a WHOIS response. Of an
+// answer from RIPE, APNIC or AFRINIC it reads only the lines asObjectLines
+// returns. Of an answer from ARIN or LACNIC it reads every line.
 func (c *Client) parseResponse(asn int, rir, response string) *ASNInfo {
 	info := &ASNInfo{
 		ASN:         asn,
@@ -211,84 +443,26 @@ func (c *Client) parseResponse(asn int, rir, response string) *ASNInfo {
 	}
 
 	lines := strings.Split(response, "\n")
+	if rir == RIRRIPE || rir == RIRAPNIC || rir == RIRAFRNIC {
+		lines = asObjectLines(lines)
+	}
+
 	var addressLines []string
 
 	for _, line := range lines {
-		line = strings.TrimSpace(line)
+		key, value, ok := splitField(line)
-		if line == "" || strings.HasPrefix(line, "%") || strings.HasPrefix(line, "#") {
+		if !ok || value == "" {
 			continue
 		}
 
-		// Split on first colon
-		parts := strings.SplitN(line, ":", keyValueParts)
-		if len(parts) != keyValueParts {
-			continue
-		}
-
-		key := strings.TrimSpace(strings.ToLower(parts[0]))
-		value := strings.TrimSpace(parts[1])
-
-		if value == "" {
-			continue
-		}
-
-		switch key {
-		// AS Name (varies by RIR)
-		case "asname", "as-name":
-			if info.ASName == "" {
-				info.ASName = value
-			}
-
-		// Organization
-		case "orgname", "org-name", "owner":
-			if info.OrgName == "" {
-				info.OrgName = value
-			}
-		case "orgid", "org-id", "org":
-			if info.OrgID == "" {
-				info.OrgID = value
-			}
-
 		// Address (collect multiple lines)
-		case "address":
+		if key == "address" {
 			addressLines = append(addressLines, value)
 
-		// Country
+			continue
-		case "country":
-			if info.CountryCode == "" && len(value) == 2 {
-				info.CountryCode = strings.ToUpper(value)
 		}
 
-		// Abuse contact
+		c.setResponseField(info, key, value)
-		case "orgabuseemail", "abuse-mailbox":
-			if info.AbuseEmail == "" {
-				info.AbuseEmail = value
-			}
-		case "orgabusephone":
-			if info.AbusePhone == "" {
-				info.AbusePhone = value
-			}
-
-		// Tech contact
-		case "orgtechemail":
-			if info.TechEmail == "" {
-				info.TechEmail = value
-			}
-		case "orgtechphone":
-			if info.TechPhone == "" {
-				info.TechPhone = value
-			}
-
-		// Registration dates
-		case "regdate", "created":
-			if info.RegDate == nil {
-				info.RegDate = c.parseDate(value)
-			}
-		case "updated", "last-modified", "changed":
-			if info.LastMod == nil {
-				info.LastMod = c.parseDate(value)
-			}
-		}
 	}
 
 	// Combine address lines
@@ -304,6 +478,131 @@ func (c *Client) parseResponse(asn int, rir, response string) *ASNInfo {
 	return info
 }
 
+// splitField splits a "key: value" line of a WHOIS answer into its key, in
+// lower case, and its value. The third result is false for a line that
+// holds no field: a blank line, a comment line or a line without a colon.
+func splitField(line string) (string, string, bool) {
+	line = strings.TrimSpace(line)
+	if strings.HasPrefix(line, "%") || strings.HasPrefix(line, "#") {
+		return "", "", false
+	}
+
+	key, value, ok := strings.Cut(line, ":")
+
+	return strings.ToLower(strings.TrimSpace(key)), strings.TrimSpace(value), ok
+}
+
+// asObjectLines returns the lines of the aut-num object in an answer from
+// RIPE, APNIC or AFRINIC, followed by the lines of the organisation object
+// whose ID the aut-num object gives on its "org:" line. The answer's other
+// objects are left out. RIPE and APNIC start their answer with the
+// as-block object of the block the AS is in, whose dates are not the AS's.
+//
+// These answers are in RPSL. A blank line ends an object. An object's first
+// line has the object's class as its key and the object's ID as its value,
+// as in "organisation: ORG-APNI1-AP".
+func asObjectLines(lines []string) []string {
+	var autNum []string
+
+	// The lines of each organisation object, by its ID.
+	organisations := make(map[string][]string)
+
+	// The class and ID of the object the line is in, "" between objects.
+	var class, id string
+
+	// The ID of the organisation object the aut-num object refers to.
+	var orgID string
+
+	for _, line := range lines {
+		if strings.TrimSpace(line) == "" {
+			class = ""
+
+			continue
+		}
+
+		key, value, ok := splitField(line)
+		if !ok {
+			continue
+		}
+
+		if class == "" {
+			class, id = key, value
+		}
+
+		switch class {
+		case autNumClass:
+			autNum = append(autNum, line)
+
+			if key == orgKey {
+				orgID = value
+			}
+		case organisationClass:
+			organisations[id] = append(organisations[id], line)
+		}
+	}
+
+	return append(autNum, organisations[orgID]...)
+}
+
+// setResponseField stores the value of one "key: value" line of a WHOIS
+// response in info, keeping the first value seen for each field.
+func (c *Client) setResponseField(info *ASNInfo, key, value string) {
+	switch key {
+	// Country
+	case "country":
+		if info.CountryCode == "" && len(value) == 2 {
+			info.CountryCode = strings.ToUpper(value)
+		}
+
+	// Registration dates
+	case "regdate", "created":
+		if info.RegDate == nil {
+			info.RegDate = c.parseDate(value)
+		}
+	case "updated", "last-modified", "changed":
+		if info.LastMod == nil {
+			info.LastMod = c.parseDate(value)
+		}
+
+	default:
+		field := textFieldForKey(info, key)
+		if field != nil && *field == "" {
+			*field = value
+		}
+	}
+}
+
+// textFieldForKey returns the text field of info that a WHOIS response key
+// fills, or nil when the key fills none.
+func textFieldForKey(info *ASNInfo, key string) *string {
+	switch key {
+	// AS Name (varies by RIR)
+	case "asname", "as-name":
+		return &info.ASName
+
+	// Organization
+	case "orgname", "org-name", "owner":
+		return &info.OrgName
+	case "orgid", "org-id", "org":
+		return &info.OrgID
+
+	// Abuse contact
+	case "orgabuseemail", "abuse-mailbox":
+		return &info.AbuseEmail
+	case "orgabusephone":
+		return &info.AbusePhone
+
+	// Tech contact
+	case "orgtechemail":
+		return &info.TechEmail
+	case "orgtechphone":
+		return &info.TechPhone
+
+	default:
+		return nil
+	}
+}
+
 // parseDate attempts to parse various date formats used in WHOIS responses.
 func (c *Client) parseDate(value string) *time.Time {
 	// Common formats
@@ -319,15 +618,15 @@ func (c *Client) parseDate(value string) *time.Time {
 	value = strings.TrimSpace(value)
 	// Handle "YYYYMMDD" format from LACNIC
 	if len(value) == lacnicDateFormatLen {
-		if _, err := time.Parse("20060102", value); err == nil {
+		t, err := time.Parse("20060102", value)
-			t, _ := time.Parse("20060102", value)
+		if err == nil {
-
 			return &t
 		}
 	}
 
 	for _, format := range formats {
-		if t, err := time.Parse(format, value); err == nil {
+		t, err := time.Parse(format, value)
+		if err == nil {
 			return &t
 		}
 	}

internal/whois/whois_test.go

@@ -0,0 +1,881 @@
+package whois_test
+
+import (
+	"bufio"
+	"bytes"
+	"context"
+	"encoding/json"
+	"errors"
+	"io"
+	"log/slog"
+	"maps"
+	"net"
+	"slices"
+	"strings"
+	"sync"
+	"syscall"
+	"testing"
+	"time"
+	"unicode/utf8"
+
+	"git.eeqj.de/sneak/routewatch/internal/whois"
+)
+
+// The servers LookupASN asks: IANA first, then the RIR IANA refers it to.
+const (
+	ianaServer = "whois.iana.org:43"
+	ripeServer = "whois.ripe.net:43"
+)
+
+// testASN is the AS the tests look up, and asQuery the query for it.
+const (
+	testASN = 3333
+	asQuery = "AS3333"
+)
+
+// The fake servers' answers are keyed by a server's address, a space and a
+// query. These are the keys of the queries LookupASN sends for testASN:
+// IANA's, then RIPE's, which asks RIPE to leave out person and role
+// objects.
+const (
+	ianaQuery = ianaServer + " " + asQuery
+	ripeQuery = ripeServer + " --no-personal " + asQuery
+)
+
+// ianaReferral is IANA's answer for an AS in a block it gave to RIPE.
+const ianaReferral = `% IANA WHOIS server
+
+refer:        whois.ripe.net
+
+as-block:     3154-3353
+organisation: Assigned by RIPE NCC
+whois:        whois.ripe.net
+changed:      2002-09-01
+source:       IANA
+`
+
+// lacnicReferral is IANA's answer for an AS in a block it gave to LACNIC,
+// shortened to the referral.
+const lacnicReferral = "refer:        whois.lacnic.net\n"
+
+// ripeRecord is RIPE's answer for the AS when asked to leave out person and
+// role objects. It starts with the as-block object of the block the AS is
+// in. Then come the abuse address as a comment, the AS's object and the
+// organisation object it refers to.
+const ripeRecord = `% This is the RIPE Database query service.
+% The objects are in RPSL format.
+
+% Note: --no-personal means ALL personal data has been filtered from this output.
+
+% Information related to 'AS3209 - AS3353'
+
+as-block:       AS3209 - AS3353
+descr:          RIPE NCC ASN block
+remarks:        These AS Numbers are assigned to network operators
+remarks:        in the RIPE NCC service region.
+mnt-by:         RIPE-NCC-HM-MNT
+created:        2018-11-22T15:27:19Z
+last-modified:  2018-11-22T15:27:19Z
+source:         RIPE
+
+% Information related to 'AS3333'
+
+% Abuse contact for 'AS3333' is 'abuse@ripe.net'
+
+aut-num:        AS3333
+as-name:        RIPE-NCC-AS
+org:            ORG-RIEN1-RIPE
+admin-c:        BRD-RIPE
+tech-c:         OPS4-RIPE
+abuse-c:        OPS4-RIPE
+mnt-by:         RIPE-NCC-MNT
+created:        2002-09-17T13:08:32Z
+last-modified:  2024-01-09T08:55:41Z
+source:         RIPE
+
+organisation:   ORG-RIEN1-RIPE
+org-name:       Reseaux IP Europeens Network Coordination Centre (RIPE NCC)
+country:        NL
+org-type:       RIR
+address:        P.O. Box 10096
+address:        1001EB Amsterdam
+abuse-c:        OPS4-RIPE
+created:        2012-03-09T13:08:29Z
+last-modified:  2024-03-26T10:55:38Z
+source:         RIPE
+`
+
+// The error lines with which RIPE refuses a query: access denied to the
+// host, and the host's limit on returned objects reached.
+const (
+	ripeAccessDenied = "%ERROR:201: access denied for 192.0.2.1"
+	ripeLimitReached = "%ERROR:202: access control limit reached"
+)
+
+// ripeRefusal is RIPE's answer to a host it no longer serves.
+const ripeRefusal = `% This is the RIPE Database query service.
+% The objects are in RPSL format.
+%
+` + ripeAccessDenied + `
+%
+% Sorry, access from your host has been permanently
+% denied because of a repeated excessive querying.
+`
+
+// apnicAfrinicRecord is the answer of APNIC or AFRINIC for the AS when asked
+// to leave out person and role objects. Like RIPE's, it holds the abuse
+// address as a comment, the AS's object and the organisation object.
+const apnicAfrinicRecord = `% Abuse contact for 'AS3333' is 'abuse@example.net'
+
+aut-num:        AS3333
+as-name:        EXAMPLE-AS
+org:            ORG-EXAMPLE1
+admin-c:        EX1-TEST
+last-modified:  2026-08-26T00:40:00Z
+
+organisation:   ORG-EXAMPLE1
+org-name:       Example Networks
+country:        ZA
+address:        1 Example Road
+`
+
+// apnicServer is APNIC's WHOIS server, and apnicASN an AS that APNIC holds.
+const (
+	apnicServer = "whois.apnic.net:43"
+	apnicASN    = 4608
+)
+
+// The keys of the queries LookupASN sends for apnicASN: IANA's, then
+// APNIC's, which asks APNIC to leave out person and role objects.
+const (
+	apnicIANAQuery = ianaServer + " AS4608"
+	apnicQuery     = apnicServer + " --no-personal AS4608"
+)
+
+// apnicRecord is APNIC's answer for apnicASN when asked to leave out person
+// and role objects, shortened. It starts with the as-block object of the
+// block the AS is in. Then come the abuse address as a comment, the AS's
+// object, and the irt and organisation objects it refers to.
+const apnicRecord = `% [whois.apnic.net]
+% Whois data copyright terms    http://www.apnic.net/db/dbcopyright.html
+
+% Note: --no-personal means ALL personal data has been filtered from this output.
+
+% Information related to 'AS4608 - AS4865'
+
+as-block:       AS4608 - AS4865
+descr:          APNIC ASN block
+remarks:        These AS numbers are further assigned by APNIC
+remarks:        to APNIC members and end-users in the APNIC region
+mnt-by:         APNIC-HM
+mnt-lower:      APNIC-HM
+last-modified:  2019-05-20T03:52:26Z
+source:         APNIC
+
+% Information related to 'AS4608'
+
+% Abuse contact for 'AS4608' is 'abuse@apnic.net'
+
+aut-num:        AS4608
+as-name:        APNIC-SERVICES
+descr:          Asia Pacific Network Information Centre
+country:        AU
+org:            ORG-APNI1-AP
+admin-c:        AIC1-AP
+abuse-c:        AA1589-AP
+mnt-by:         APNIC-HM
+mnt-irt:        IRT-APNIC-IS-AP
+last-modified:  2026-08-26T00:40:00Z
+source:         APNIC
+
+irt:            IRT-APNIC-IS-AP
+address:        South Brisbane, Australia
+e-mail:         helpdesk@apnic.net
+abuse-mailbox:  abuse@apnic.net
+mnt-by:         MAINT-APNIC-IS-AP
+last-modified:  2025-11-18T00:27:08Z
+source:         APNIC
+
+organisation:   ORG-APNI1-AP
+org-name:       Asia Pacific Network Information Centre
+org-type:       LIR
+country:        AU
+address:        PO Box 3646
+address:        South Brisbane, QLD 4101,
+address:        Australia
+e-mail:         noc@apnic.net
+mnt-by:         APNIC-HM
+last-modified:  2026-08-24T12:58:43Z
+source:         APNIC
+
+% This query was served by the APNIC Whois Service version 1.88.48 (WHOIS-UK2)
+`
+
+// ianaPrivateUse is IANA's answer for a private-use AS: its block and no
+// referral.
+const ianaPrivateUse = `% IANA WHOIS server
+
+as-block:     64512-65534
+organisation: Reserved for Private Use
+descr:        Reserved for Private Use
+source:       IANA
+`
+
+// ianaNoObject is IANA's answer when it has nothing for the query.
+const ianaNoObject = `% IANA WHOIS server
+% This query returned 0 objects.
+`
+
+// fillerLine is one of the lines a server that does not stop keeps sending.
+const fillerLine = "remarks:        and more\n"
+
+// endlessAnswerSize is how many bytes of filler lines that server sends:
+// four times the client's cap on an answer.
+const endlessAnswerSize = 4 << 20
+
+// fakeServers returns a DialContext for a whois.Client that connects to a
+// fake server at each address answers names. Each key of answers is a
+// server's address, a space and a query, and its value is that server's
+// answer to that query. A server reads one query and sends its answer, or
+// nothing for a query it has no answer for. Connecting to any other address
+// is refused.
+func fakeServers(
+	t *testing.T, answers map[string]string,
+) func(context.Context, string, string) (net.Conn, error) {
+	t.Helper()
+
+	addresses := make(map[string]bool)
+
+	for key := range answers {
+		address, _, _ := strings.Cut(key, " ")
+		addresses[address] = true
+	}
+
+	var servers sync.WaitGroup
+
+	t.Cleanup(servers.Wait)
+
+	return func(_ context.Context, _, address string) (net.Conn, error) {
+		if !addresses[address] {
+			return nil, syscall.ECONNREFUSED
+		}
+
+		client, server := net.Pipe()
+
+		servers.Go(func() {
+			defer func() { _ = server.Close() }()
+
+			query, _ := bufio.NewReader(server).ReadString('\n')
+			answer := answers[address+" "+strings.TrimSpace(query)]
+			_, _ = io.WriteString(server, answer)
+		})
+
+		return client, nil
+	}
+}
+
+// lookUp looks testASN up with fake servers sending answers.
+func lookUp(
+	t *testing.T, answers map[string]string,
+) (*whois.ASNInfo, error) {
+	t.Helper()
+
+	client := whois.NewClient(slog.New(slog.DiscardHandler))
+	client.DialContext = fakeServers(t, answers)
+
+	return client.LookupASN(t.Context(), testASN)
+}
+
+// TestLookupASNReadsTheRIRAnswer looks up an AS IANA refers to RIPE. RIPE
+// answers only a query that leaves out person and role objects, and its
+// answer to that still fills the AS's fields: the abuse address from the
+// comment line, and the organisation's name and country from its object.
+// The dates are the AS's. IANA's "changed:" line dates the AS's block, and
+// so do the dates of the as-block object at the start of RIPE's answer.
+func TestLookupASNReadsTheRIRAnswer(t *testing.T) {
+	t.Parallel()
+
+	info, err := lookUp(t, map[string]string{
+		ianaQuery: ianaReferral,
+		ripeQuery: ripeRecord,
+	})
+	if err != nil {
+		t.Fatalf("LookupASN: %v", err)
+	}
+
+	got := []string{
+		info.ASName, info.OrgID, info.OrgName, info.CountryCode,
+		info.AbuseEmail, info.RIR,
+	}
+	want := []string{
+		"RIPE-NCC-AS", "ORG-RIEN1-RIPE",
+		"Reseaux IP Europeens Network Coordination Centre (RIPE NCC)", "NL",
+		"abuse@ripe.net", whois.RIRRIPE,
+	}
+
+	if !slices.Equal(got, want) {
+		t.Errorf("LookupASN = AS name, org ID, org name, country, abuse "+
+			"address, RIR %q; want %q", got, want)
+	}
+
+	wantRegDate := time.Date(2002, 9, 17, 13, 8, 32, 0, time.UTC)
+	if info.RegDate == nil || !info.RegDate.Equal(wantRegDate) {
+		t.Errorf("LookupASN registered = %v, want %v",
+			info.RegDate, wantRegDate)
+	}
+
+	wantLastMod := time.Date(2024, 1, 9, 8, 55, 41, 0, time.UTC)
+	if info.LastMod == nil || !info.LastMod.Equal(wantLastMod) {
+		t.Errorf("LookupASN last modified = %v, want %v",
+			info.LastMod, wantLastMod)
+	}
+}
+
+// TestLookupASNReadsOnlyTheASAndItsOrganisation looks up an AS at APNIC,
+// whose answer starts with the as-block object of the block the AS is in.
+// The last-modified date is the aut-num object's, not the block's. The
+// address is the organisation object's alone, without the irt object's.
+func TestLookupASNReadsOnlyTheASAndItsOrganisation(t *testing.T) {
+	t.Parallel()
+
+	client := whois.NewClient(slog.New(slog.DiscardHandler))
+	client.DialContext = fakeServers(t, map[string]string{
+		apnicIANAQuery: "refer:        whois.apnic.net\n",
+		apnicQuery:     apnicRecord,
+	})
+
+	info, err := client.LookupASN(t.Context(), apnicASN)
+	if err != nil {
+		t.Fatalf("LookupASN: %v", err)
+	}
+
+	wantLastMod := time.Date(2026, 8, 26, 0, 40, 0, 0, time.UTC)
+	if info.LastMod == nil || !info.LastMod.Equal(wantLastMod) {
+		t.Errorf("LookupASN last modified = %v, want %v",
+			info.LastMod, wantLastMod)
+	}
+
+	got := []string{
+		info.ASName, info.OrgID, info.OrgName, info.CountryCode,
+		info.Address, info.AbuseEmail,
+	}
+	want := []string{
+		"APNIC-SERVICES", "ORG-APNI1-AP",
+		"Asia Pacific Network Information Centre", "AU",
+		"PO Box 3646\nSouth Brisbane, QLD 4101,\nAustralia", "abuse@apnic.net",
+	}
+
+	if !slices.Equal(got, want) {
+		t.Errorf("LookupASN = AS name, org ID, org name, country, address, "+
+			"abuse address %q; want %q", got, want)
+	}
+}
+
+// TestLookupASNAsksAPNICAndAFRINICToLeaveOutPersonalData looks up an AS at
+// APNIC and at AFRINIC. Each answers only a query that leaves out person
+// and role objects, as RIPE does. That one answer still holds the
+// organisation object, so the organisation's name and country are filled.
+func TestLookupASNAsksAPNICAndAFRINICToLeaveOutPersonalData(t *testing.T) {
+	t.Parallel()
+
+	for _, server := range []string{"whois.apnic.net", "whois.afrinic.net"} {
+		t.Run(server, func(t *testing.T) {
+			t.Parallel()
+
+			address := server + ":43"
+
+			info, err := lookUp(t, map[string]string{
+				ianaQuery:                             "refer:        " + server + "\n",
+				address + " --no-personal " + asQuery: apnicAfrinicRecord,
+			})
+			if err != nil {
+				t.Fatalf("LookupASN: %v", err)
+			}
+
+			got := []string{
+				info.ASName, info.OrgName, info.CountryCode, info.AbuseEmail,
+			}
+			want := []string{
+				"EXAMPLE-AS", "Example Networks", "ZA", "abuse@example.net",
+			}
+
+			if !slices.Equal(got, want) {
+				t.Errorf("LookupASN = AS name, org name, country, abuse "+
+					"address %q; want %q", got, want)
+			}
+		})
+	}
+}
+
+// The name of an organisation with accented letters, in Latin-1, one byte for
+// each letter, and in UTF-8.
+const (
+	latin1OrgName = "Fundaci\xf3n Ra\xedces de Espa\xf1a"
+	utf8OrgName   = "Fundación Raíces de España"
+)
+
+// TestLookupASNStoresAnswersInUTF8 looks up an AS at RIPE, which answers in
+// Latin-1, and at LACNIC, which answers in UTF-8. Both answers give the
+// organisation's name with its accented letters in UTF-8, in the parsed field
+// and in the stored answer. Before the fix the Latin-1 answer was kept as
+// received, which is not valid UTF-8.
+func TestLookupASNStoresAnswersInUTF8(t *testing.T) {
+	t.Parallel()
+
+	for name, answers := range map[string]map[string]string{
+		"RIPE in Latin-1": {
+			ianaQuery: ianaReferral,
+			ripeQuery: "aut-num:        AS3333\n" +
+				"org-name:       " + latin1OrgName + "\n",
+		},
+		"LACNIC in UTF-8": {
+			ianaQuery: lacnicReferral,
+			"whois.lacnic.net:43 " + asQuery: "aut-num:     AS3333\n" +
+				"owner:       " + utf8OrgName + "\n",
+		},
+	} {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			info, err := lookUp(t, answers)
+			if err != nil {
+				t.Fatalf("LookupASN: %v", err)
+			}
+
+			if info.OrgName != utf8OrgName {
+				t.Errorf("LookupASN org name = %q, want %q",
+					info.OrgName, utf8OrgName)
+			}
+
+			if !utf8.ValidString(info.RawResponse) ||
+				!strings.Contains(info.RawResponse, utf8OrgName) {
+				t.Errorf("LookupASN answer = %q, want UTF-8 holding %q",
+					info.RawResponse, utf8OrgName)
+			}
+		})
+	}
+}
+
+// lacnicServer is LACNIC's WHOIS server and nicbrServer NIC.br's, which
+// registers Brazil's ASes under LACNIC. brazilASN is one of those ASes.
+const (
+	lacnicServer = "whois.lacnic.net:43"
+	nicbrServer  = "whois.registro.br:43"
+	brazilASN    = 52751
+)
+
+// The keys of the queries LookupASN sends for brazilASN: IANA's, LACNIC's,
+// then NIC.br's.
+const (
+	brazilIANAQuery   = ianaServer + " AS52751"
+	brazilLACNICQuery = lacnicServer + " AS52751"
+	brazilNICbrQuery  = nicbrServer + " AS52751"
+)
+
+// The name of brazilASN's organisation in three forms: as LACNIC sends it,
+// with each accented letter replaced by U+FFFD; as NIC.br sends it, in
+// Latin-1; and in UTF-8.
+const (
+	lacnicBrazilOrgName = "Brasrede Telecomunica\U0000FFFD\U0000FFFDes LTDA"
+	nicbrBrazilOrgName  = "Brasrede Telecomunica\xe7\xf5es LTDA"
+	brazilOrgName       = "Brasrede Telecomunicações LTDA"
+)
+
+// nicbrCopyright is the copyright notice at the top of NIC.br's answers.
+const nicbrCopyright = "% Copyright (c) Nic.br - " +
+	"Use of this data is governed by the Use and\n" +
+	"% Privacy Policy at https://registro.br/upp . Distribution,\n" +
+	"% commercialization, reproduction, and use for advertising or similar\n" +
+	"% purposes are expressly prohibited.\n"
+
+// lacnicBrazilRecord is LACNIC's answer for brazilASN, recorded on
+// 2026-10-26. It is NIC.br's answer, which LACNIC passes on after a line
+// of its own and a space. It is shortened. The client's address is a
+// documentation address. The contact's name, the contact's object and the
+// footer are left out.
+const lacnicBrazilRecord = "% IP Client: 192.0.2.1\n" +
+	" " + nicbrCopyright + `% 2026-10-26T23:12:22-03:00 - 192.0.2.1
+
+aut-num:     AS52751
+owner:       ` + lacnicBrazilOrgName + `
+owner-c:     AVL217
+routing-c:   AVL217
+abuse-c:     AVL217
+created:     20120829
+changed:     20130306
+inetnum:     177.38.152.0/21
+inetnum:     2804:b34::/32
+inetnum:     191.243.4.0/22
+inetnum:     131.72.248.0/22
+inetnum:     138.94.228.0/22
+inetnum:     170.233.8.0/22
+inetnum:     170.246.4.0/22
+`
+
+// nicbrRecord is NIC.br's answer for brazilASN, in Latin-1, recorded on
+// 2026-10-26 and shortened as lacnicBrazilRecord is.
+const nicbrRecord = nicbrCopyright + `% 2026-10-26T23:14:41-03:00 - 192.0.2.1
+
+aut-num:     AS52751
+owner:       ` + nicbrBrazilOrgName + `
+owner-c:     AVL217
+routing-c:   AVL217
+abuse-c:     AVL217
+created:     20120829
+changed:     20130306
+inetnum:     177.38.152.0/21
+inetnum:     2804:b34::/32
+inetnum:     191.243.4.0/22
+inetnum:     131.72.248.0/22
+inetnum:     138.94.228.0/22
+inetnum:     170.233.8.0/22
+inetnum:     170.246.4.0/22
+`
+
+// nicbrOverLimit stands for NIC.br's answer for brazilASN to a client over
+// its query limit, and is not recorded. Such an answer leaves information
+// out, here the owner, and says so in its first line.
+const nicbrOverLimit = `% Query rate limit exceeded. Reduced information.
+
+aut-num:     AS52751
+created:     20120829
+changed:     20130306
+`
+
+// newNICbrClient returns a client whose fake servers send answers and
+// whose logger writes JSON lines to logs. It adds one to *nicbrAsks each
+// time it connects to NIC.br's server.
+func newNICbrClient(
+	t *testing.T, answers map[string]string, logs io.Writer, nicbrAsks *int,
+) *whois.Client {
+	t.Helper()
+
+	dial := fakeServers(t, answers)
+
+	client := whois.NewClient(slog.New(slog.NewJSONHandler(logs, nil)))
+	client.DialContext = func(
+		ctx context.Context, network, address string,
+	) (net.Conn, error) {
+		if address == nicbrServer {
+			*nicbrAsks++
+		}
+
+		return dial(ctx, network, address)
+	}
+
+	return client
+}
+
+// lookUpBrazilianAS looks brazilASN up with fake servers sending answers,
+// and reports whether the client connected to NIC.br's server.
+func lookUpBrazilianAS(
+	t *testing.T, answers map[string]string,
+) (*whois.ASNInfo, bool, error) {
+	t.Helper()
+
+	nicbrAsks := 0
+	client := newNICbrClient(t, answers, io.Discard, &nicbrAsks)
+
+	info, err := client.LookupASN(t.Context(), brazilASN)
+
+	return info, nicbrAsks > 0, err
+}
+
+// TestLookupASNTakesABrazilianASFromNICbr looks up an AS that NIC.br
+// registers. LACNIC's answer is NIC.br's with each accented letter replaced
+// by U+FFFD, so the client also asks NIC.br's server, which answers in
+// Latin-1. The organisation's name has its accented letters, in the parsed
+// field and in the stored answer.
+func TestLookupASNTakesABrazilianASFromNICbr(t *testing.T) {
+	t.Parallel()
+
+	info, _, err := lookUpBrazilianAS(t, map[string]string{
+		brazilIANAQuery:   lacnicReferral,
+		brazilLACNICQuery: lacnicBrazilRecord,
+		brazilNICbrQuery:  nicbrRecord,
+	})
+	if err != nil {
+		t.Fatalf("LookupASN: %v", err)
+	}
+
+	if info.OrgName != brazilOrgName || info.RIR != whois.RIRLACNIC {
+		t.Errorf("LookupASN org name, RIR = %q, %q; want %q, %q",
+			info.OrgName, info.RIR, brazilOrgName, whois.RIRLACNIC)
+	}
+
+	if !strings.Contains(info.RawResponse, brazilOrgName) {
+		t.Errorf("LookupASN answer = %q, want one holding %q",
+			info.RawResponse, brazilOrgName)
+	}
+
+	wantRegDate := time.Date(2012, 8, 29, 0, 0, 0, 0, time.UTC)
+	if info.RegDate == nil || !info.RegDate.Equal(wantRegDate) {
+		t.Errorf("LookupASN registered = %v, want %v",
+			info.RegDate, wantRegDate)
+	}
+}
+
+// TestLookupASNAsksNICbrOnlyForItsASes looks up an AS at LACNIC whose
+// answer does not carry NIC.br's copyright notice, as for an AS that LACNIC
+// registers itself. The client does not connect to NIC.br's server, which
+// would answer with a record, and keeps LACNIC's fields.
+func TestLookupASNAsksNICbrOnlyForItsASes(t *testing.T) {
+	t.Parallel()
+
+	info, askedNICbr, err := lookUpBrazilianAS(t, map[string]string{
+		brazilIANAQuery: lacnicReferral,
+		brazilLACNICQuery: "aut-num:     AS52751\n" +
+			"owner:       " + utf8OrgName + "\n",
+		brazilNICbrQuery: nicbrRecord,
+	})
+	if err != nil {
+		t.Fatalf("LookupASN: %v", err)
+	}
+
+	if askedNICbr {
+		t.Error("LookupASN asked NIC.br's server about an AS whose " +
+			"answer from LACNIC is not NIC.br's")
+	}
+
+	if info.OrgName != utf8OrgName {
+		t.Errorf("LookupASN org name = %q, want LACNIC's %q",
+			info.OrgName, utf8OrgName)
+	}
+}
+
+// TestLookupASNKeepsLACNICsAnswerWhenNICbrFails looks up an AS that NIC.br
+// registers when NIC.br's server refuses the connection, sends nothing, or
+// sends the reduced answer for a client over its query limit. The client
+// asks NIC.br, and the lookup still succeeds with LACNIC's fields. The
+// client logs one warning, which names the AS and the reason.
+func TestLookupASNKeepsLACNICsAnswerWhenNICbrFails(t *testing.T) {
+	t.Parallel()
+
+	for name, test := range map[string]struct {
+		nicbrAnswer map[string]string
+		reason      string
+	}{
+		"connection refused": {
+			nicbrAnswer: map[string]string{},
+			reason:      syscall.ECONNREFUSED.Error(),
+		},
+		"no answer": {
+			nicbrAnswer: map[string]string{brazilNICbrQuery: ""},
+			reason:      "no WHOIS fields",
+		},
+		"over the query limit": {
+			nicbrAnswer: map[string]string{brazilNICbrQuery: nicbrOverLimit},
+			reason:      "over its query limit",
+		},
+	} {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			answers := map[string]string{
+				brazilIANAQuery:   lacnicReferral,
+				brazilLACNICQuery: lacnicBrazilRecord,
+			}
+			maps.Copy(answers, test.nicbrAnswer)
+
+			var logs bytes.Buffer
+
+			nicbrAsks := 0
+			client := newNICbrClient(t, answers, &logs, &nicbrAsks)
+
+			info, err := client.LookupASN(t.Context(), brazilASN)
+			if err != nil {
+				t.Fatalf("LookupASN: %v", err)
+			}
+
+			if nicbrAsks == 0 {
+				t.Error("LookupASN did not ask NIC.br's server")
+			}
+
+			if info.OrgName != lacnicBrazilOrgName {
+				t.Errorf("LookupASN org name = %q, want LACNIC's %q",
+					info.OrgName, lacnicBrazilOrgName)
+			}
+
+			var logged struct {
+				Level string `json:"level"`
+				ASN   int    `json:"asn"`
+			}
+
+			err = json.Unmarshal(logs.Bytes(), &logged)
+			if err != nil {
+				t.Fatalf("want one log line, got %q: %v", logs.String(), err)
+			}
+
+			if logged.Level != slog.LevelWarn.String() ||
+				logged.ASN != brazilASN ||
+				!strings.Contains(logs.String(), test.reason) {
+				t.Errorf("logged %q, want a warning for AS%d holding %q",
+					logs.String(), brazilASN, test.reason)
+			}
+		})
+	}
+}
+
+// TestLookupASNSkipsNICbrAfterItsLimit looks up an AS that NIC.br
+// registers three times with one client, while NIC.br answers that the
+// client is over its query limit. The first lookup asks NIC.br. The
+// second, a second before NICbrSkipTime has passed, does not. The third,
+// once NICbrSkipTime has passed, asks again. Each keeps LACNIC's fields.
+func TestLookupASNSkipsNICbrAfterItsLimit(t *testing.T) {
+	t.Parallel()
+
+	nicbrAsks := 0
+	client := newNICbrClient(t, map[string]string{
+		brazilIANAQuery:   lacnicReferral,
+		brazilLACNICQuery: lacnicBrazilRecord,
+		brazilNICbrQuery:  nicbrOverLimit,
+	}, io.Discard, &nicbrAsks)
+
+	start := time.Date(2026, 10, 27, 0, 0, 0, 0, time.UTC)
+	now := start
+	client.Now = func() time.Time { return now }
+
+	for _, step := range []struct {
+		after    time.Duration
+		wantAsks int
+	}{
+		{after: 0, wantAsks: 1},
+		{after: whois.NICbrSkipTime - time.Second, wantAsks: 1},
+		{after: whois.NICbrSkipTime, wantAsks: 2},
+	} {
+		now = start.Add(step.after)
+
+		info, err := client.LookupASN(t.Context(), brazilASN)
+		if err != nil {
+			t.Fatalf("LookupASN after %v: %v", step.after, err)
+		}
+
+		if nicbrAsks != step.wantAsks {
+			t.Errorf("after %v NIC.br's server was asked %d times, want %d",
+				step.after, nicbrAsks, step.wantAsks)
+		}
+
+		if info.OrgName != lacnicBrazilOrgName {
+			t.Errorf("LookupASN after %v org name = %q, want LACNIC's %q",
+				step.after, info.OrgName, lacnicBrazilOrgName)
+		}
+	}
+}
+
+// TestLookupASNReportsARefusal looks up an AS IANA refers to RIPE when RIPE
+// refuses the query, with each of the error lines RIPE refuses with. The
+// error holds RIPE's line, so the log says why the lookup failed.
+func TestLookupASNReportsARefusal(t *testing.T) {
+	t.Parallel()
+
+	for _, line := range []string{ripeAccessDenied, ripeLimitReached} {
+		t.Run(line, func(t *testing.T) {
+			t.Parallel()
+
+			_, err := lookUp(t, map[string]string{
+				ianaQuery: ianaReferral,
+				ripeQuery: "% This is the RIPE Database query service.\n" +
+					line + "\n",
+			})
+			if err == nil || !strings.Contains(err.Error(), line) {
+				t.Errorf("LookupASN error = %v, want one holding %q",
+					err, line)
+			}
+		})
+	}
+}
+
+// TestLookupASNFailsWhenTheRIRGivesNoAnswer looks up an AS IANA refers to
+// RIPE when RIPE's server refuses the connection, and when it answers only
+// with an error. Neither is a lookup that found the AS.
+func TestLookupASNFailsWhenTheRIRGivesNoAnswer(t *testing.T) {
+	t.Parallel()
+
+	for name, answers := range map[string]map[string]string{
+		"connection refused": {ianaQuery: ianaReferral},
+		"only an error":      {ianaQuery: ianaReferral, ripeQuery: ripeRefusal},
+	} {
+		t.Run(name, func(t *testing.T) {
+			t.Parallel()
+
+			_, err := lookUp(t, answers)
+			if err == nil || errors.Is(err, whois.ErrNoReferral) {
+				t.Errorf("LookupASN error = %v, want one that is not "+
+					"ErrNoReferral", err)
+			}
+		})
+	}
+}
+
+// TestLookupASNReportsNoReferral looks up a private-use AS, which IANA
+// refers to no RIR.
+func TestLookupASNReportsNoReferral(t *testing.T) {
+	t.Parallel()
+
+	_, err := lookUp(t, map[string]string{ianaQuery: ianaPrivateUse})
+	if !errors.Is(err, whois.ErrNoReferral) {
+		t.Errorf("LookupASN error = %v, want ErrNoReferral", err)
+	}
+}
+
+// TestLookupASNFailsWhenIANAAnswersNoBlock looks up an AS when IANA answers
+// with neither a referral nor the AS's block. That is not IANA saying no
+// RIR holds the AS.
+func TestLookupASNFailsWhenIANAAnswersNoBlock(t *testing.T) {
+	t.Parallel()
+
+	_, err := lookUp(t, map[string]string{ianaQuery: ianaNoObject})
+	if err == nil || errors.Is(err, whois.ErrNoReferral) {
+		t.Errorf("LookupASN error = %v, want one that is not ErrNoReferral",
+			err)
+	}
+}
+
+// TestLookupASNFailsWhenTheRIRAnswerDoesNotEnd looks up an AS whose RIR
+// server follows the record with filler lines far past any real answer.
+// The client stops reading at its cap and fails instead of keeping them.
+func TestLookupASNFailsWhenTheRIRAnswerDoesNotEnd(t *testing.T) {
+	t.Parallel()
+
+	endless := ripeRecord +
+		strings.Repeat(fillerLine, endlessAnswerSize/len(fillerLine))
+
+	_, err := lookUp(t, map[string]string{
+		ianaQuery: ianaReferral,
+		ripeQuery: endless,
+	})
+	if err == nil || errors.Is(err, whois.ErrNoReferral) {
+		t.Errorf("LookupASN error = %v, want one that is not ErrNoReferral",
+			err)
+	}
+}
+
+// TestLookupASNFailsOnAReferralToAnotherServer looks up an AS when IANA's
+// answer refers it to a server none of the five RIRs run, with a port and
+// without. The client follows a referral only to an RIR's server, so it
+// fails instead of asking that server, which would answer with a record.
+func TestLookupASNFailsOnAReferralToAnotherServer(t *testing.T) {
+	t.Parallel()
+
+	// Each case maps the server on the "refer:" line to the address the
+	// client would dial to follow it. IANA's answer also names the AS's
+	// block, so a client that dropped the referral would report no RIR.
+	for refer, address := range map[string]string{
+		"whois.example.net:8080": "whois.example.net:8080",
+		"whois.example.net":      "whois.example.net:43",
+	} {
+		t.Run(refer, func(t *testing.T) {
+			t.Parallel()
+
+			_, err := lookUp(t, map[string]string{
+				ianaQuery: "refer:        " + refer + "\n\n" +
+					"as-block:     3154-3353\n",
+				address + " " + asQuery: ripeRecord,
+			})
+			if err == nil || errors.Is(err, whois.ErrNoReferral) {
+				t.Errorf("LookupASN error = %v, want one that is not "+
+					"ErrNoReferral", err)
+			}
+		})
+	}
+}

package.json

@@ -0,0 +1,6 @@
+{
+  "private": true,
+  "devDependencies": {
+    "prettier": "3.8.1"
+  }
+}

pkg/asinfo/asinfo.go

@@ -1,12 +1,14 @@
-// Package asinfo provides information about Autonomous Systems (AS)
 package asinfo
 
 import (
 	"bytes"
+	"cmp"
 	"compress/gzip"
 	_ "embed"
 	"encoding/json"
 	"io"
+	"slices"
+	"strings"
 	"sync"
 )
 
@@ -24,6 +26,17 @@ type Info struct {
 type Registry struct {
 	mu    sync.RWMutex
 	byASN map[int]*Info
+	// searchEntries holds every entry in AS number order, for Search.
+	searchEntries []searchEntry
+}
+
+// searchEntry is an entry with its handle and description in lower case. They
+// are put in lower case once, when the data loads, so that Search only
+// compares.
+type searchEntry struct {
+	info             *Info
+	lowerHandle      string
+	lowerDescription string
 }
 
 var (
@@ -54,13 +67,27 @@ func initRegistry() {
 	}
 
 	var asInfos []Info
-	if err := json.Unmarshal(jsonData, &asInfos); err != nil {
+
+	err = json.Unmarshal(jsonData, &asInfos)
+	if err != nil {
 		panic("failed to unmarshal embedded AS data: " + err.Error())
 	}
 
+	slices.SortFunc(asInfos, func(a, b Info) int {
+		return cmp.Compare(a.ASN, b.ASN)
+	})
+
+	defaultRegistry.searchEntries = make([]searchEntry, 0, len(asInfos))
+
 	for i := range asInfos {
 		info := &asInfos[i]
 		defaultRegistry.byASN[info.ASN] = info
+		defaultRegistry.searchEntries = append(defaultRegistry.searchEntries,
+			searchEntry{
+				info:             info,
+				lowerHandle:      strings.ToLower(info.Handle),
+				lowerDescription: strings.ToLower(info.Description),
+			})
 	}
 }
 
@@ -128,35 +155,36 @@ func All() []Info {
 	return result
 }
 
-// Search searches for AS entries where the handle or description contains the query
+// Search finds the AS entries whose handle or description contains query,
-// This is a simple case-sensitive substring search
+// ignoring case. It returns the first limit of them in AS number order, and the
-func Search(query string) []Info {
+// number of entries that match. An empty query matches nothing.
+func Search(query string, limit int) ([]Info, int) {
 	once.Do(initRegistry)
 
+	if query == "" {
+		return nil, 0
+	}
+
+	query = strings.ToLower(query)
+
 	defaultRegistry.mu.RLock()
 	defer defaultRegistry.mu.RUnlock()
 
-	var results []Info
+	var (
-	for _, info := range defaultRegistry.byASN {
+		results []Info
-		if contains(info.Handle, query) || contains(info.Description, query) {
+		total   int
-			results = append(results, *info)
+	)
+
+	for _, entry := range defaultRegistry.searchEntries {
+		if strings.Contains(entry.lowerHandle, query) ||
+			strings.Contains(entry.lowerDescription, query) {
+			if total < limit {
+				results = append(results, *entry.info)
+			}
+
+			total++
 		}
 	}
 
-	return results
+	return results, total
-}
-
-// contains is a simple substring search
-func contains(s, substr string) bool {
-	return len(substr) > 0 && len(s) >= len(substr) && containsImpl(s, substr)
-}
-
-func containsImpl(s, substr string) bool {
-	for i := 0; i <= len(s)-len(substr); i++ {
-		if s[i:i+len(substr)] == substr {
-			return true
-		}
-	}
-
-	return false
 }

pkg/asinfo/asinfo_test.go

@@ -1,10 +1,17 @@
-package asinfo
+package asinfo_test
 
 import (
+	"cmp"
+	"slices"
+	"strings"
 	"testing"
+
+	"git.eeqj.de/sneak/routewatch/pkg/asinfo"
 )
 
 func TestGet(t *testing.T) {
+	t.Parallel()
+
 	tests := []struct {
 		name     string
 		asn      int
@@ -15,7 +22,7 @@ func TestGet(t *testing.T) {
 			name:     "AS1 Level 3",
 			asn:      1,
 			wantOK:   true,
-			wantDesc: "Level 3 Parent, LLC",
+			wantDesc: "Level 3 Parent LLC",
 		},
 		{
 			name:     "AS15169 Google",
@@ -29,40 +36,49 @@ func TestGet(t *testing.T) {
 			wantOK: false,
 		},
 		{
-			name:     "AS0 IANA Reserved",
+			name:     "AS1251 with accented letters",
-			asn:      0,
+			asn:      1251,
 			wantOK:   true,
-			wantDesc: "Internet Assigned Numbers Authority",
+			wantDesc: "FUNDAÇÃO DE AMPARO À PESQUISA DO ESTADO SÃO PAULO",
 		},
 	}
 
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			info, ok := Get(tt.asn)
+			t.Parallel()
+
+			info, ok := asinfo.Get(tt.asn)
 			if ok != tt.wantOK {
 				t.Errorf("Get(%d) ok = %v, want %v", tt.asn, ok, tt.wantOK)
+
 				return
 			}
 
 			if ok && info.Description != tt.wantDesc {
-				t.Errorf("Get(%d) description = %q, want %q", tt.asn, info.Description, tt.wantDesc)
+				t.Errorf("Get(%d) description = %q, want %q",
+					tt.asn, info.Description, tt.wantDesc)
 			}
 		})
 	}
 }
 
 func TestGetDescription(t *testing.T) {
+	t.Parallel()
+
 	tests := []struct {
 		asn  int
 		want string
 	}{
-		{1, "Level 3 Parent, LLC"},
+		{1, "Level 3 Parent LLC"},
 		{999999999, ""},
 	}
 
 	for _, tt := range tests {
 		t.Run("", func(t *testing.T) {
-			if got := GetDescription(tt.asn); got != tt.want {
+			t.Parallel()
+
+			got := asinfo.GetDescription(tt.asn)
+			if got != tt.want {
 				t.Errorf("GetDescription(%d) = %q, want %q", tt.asn, got, tt.want)
 			}
 		})
@@ -70,18 +86,22 @@ func TestGetDescription(t *testing.T) {
 }
 
 func TestGetHandle(t *testing.T) {
+	t.Parallel()
+
 	tests := []struct {
 		asn  int
 		want string
 	}{
-		{1, "LVLT"},
+		{1, "LVLT-1"},
-		{0, "IANA-RSVD"},
 		{999999999, ""},
 	}
 
 	for _, tt := range tests {
 		t.Run("", func(t *testing.T) {
-			if got := GetHandle(tt.asn); got != tt.want {
+			t.Parallel()
+
+			got := asinfo.GetHandle(tt.asn)
+			if got != tt.want {
 				t.Errorf("GetHandle(%d) = %q, want %q", tt.asn, got, tt.want)
 			}
 		})
@@ -89,18 +109,23 @@ func TestGetHandle(t *testing.T) {
 }
 
 func TestTotal(t *testing.T) {
-	total := Total()
+	t.Parallel()
+
+	total := asinfo.Total()
 	if total < 100000 {
 		t.Errorf("Total() = %d, expected > 100000", total)
 	}
 
 	// Verify it's consistent
-	if total2 := Total(); total2 != total {
+	total2 := asinfo.Total()
+	if total2 != total {
 		t.Errorf("Total() returned different values: %d vs %d", total, total2)
 	}
 }
 
 func TestSearch(t *testing.T) {
+	t.Parallel()
+
 	tests := []struct {
 		name    string
 		query   string
@@ -117,9 +142,9 @@ func TestSearch(t *testing.T) {
 			wantMin: 10,
 		},
 		{
-			name:    "Empty search",
+			name:    "Search for São Paulo",
-			query:   "",
+			query:   "São Paulo",
-			wantMin: 0,
+			wantMin: 1,
 		},
 		{
 			name:    "Non-existent org",
@@ -130,31 +155,129 @@ func TestSearch(t *testing.T) {
 
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			results := Search(tt.query)
+			t.Parallel()
+
+			results, _ := asinfo.Search(tt.query, asinfo.Total())
 			if len(results) < tt.wantMin {
-				t.Errorf("Search(%q) returned %d results, want at least %d", tt.query, len(results), tt.wantMin)
+				t.Errorf("Search(%q) returned %d results, want at least %d",
+					tt.query, len(results), tt.wantMin)
 			}
 
-			// Verify all results contain the query
+			// Verify all results contain the query, ignoring case
-			if tt.query != "" {
+			query := strings.ToLower(tt.query)
 			for _, r := range results {
-					if !contains(r.Handle, tt.query) && !contains(r.Description, tt.query) {
+				if !strings.Contains(strings.ToLower(r.Handle), query) &&
-						t.Errorf("Search result ASN %d doesn't contain query %q", r.ASN, tt.query)
+					!strings.Contains(strings.ToLower(r.Description), query) {
-					}
+					t.Errorf("Search result ASN %d doesn't contain query %q",
+						r.ASN, tt.query)
 				}
 			}
 		})
 	}
 }
 
+// TestSearchEmptyQueryMatchesNothing expects an empty query to match no AS,
+// although every handle contains the empty string.
+func TestSearchEmptyQueryMatchesNothing(t *testing.T) {
+	t.Parallel()
+
+	results, total := asinfo.Search("", asinfo.Total())
+	if len(results) != 0 || total != 0 {
+		t.Errorf(`Search("") returned %d results and a total of %d, want none`,
+			len(results), total)
+	}
+}
+
+// TestSearchIgnoresCase searches for Cloudflare's handle in lower case and its
+// description in upper case, and expects both searches to find AS13335.
+func TestSearchIgnoresCase(t *testing.T) {
+	t.Parallel()
+
+	const cloudflareASN = 13335 // handle CLOUDFLARENET, "Cloudflare Inc."
+
+	isCloudflare := func(info asinfo.Info) bool {
+		return info.ASN == cloudflareASN
+	}
+
+	for _, query := range []string{"cloudflarenet", "CLOUDFLARE INC."} {
+		results, _ := asinfo.Search(query, asinfo.Total())
+		if !slices.ContainsFunc(results, isCloudflare) {
+			t.Errorf("Search(%q) did not find AS%d", query, cloudflareASN)
+		}
+	}
+}
+
+// TestSearchOrdersByASN expects the matches of a search with many of them to
+// come in AS number order.
+func TestSearchOrdersByASN(t *testing.T) {
+	t.Parallel()
+
+	const query = "University"
+
+	results, _ := asinfo.Search(query, asinfo.Total())
+
+	for i := 1; i < len(results); i++ {
+		if results[i-1].ASN >= results[i].ASN {
+			t.Fatalf("Search(%q): AS%d comes after AS%d",
+				query, results[i].ASN, results[i-1].ASN)
+		}
+	}
+}
+
+// TestSearchStopsAtLimit searches for a word that matches more entries than
+// the limit. It expects exactly limit entries, the first matches in AS number
+// order, and the number of all matches, counted here from All.
+func TestSearchStopsAtLimit(t *testing.T) {
+	t.Parallel()
+
+	const (
+		query = "university"
+		limit = 10
+	)
+
+	var matches []asinfo.Info
+
+	for _, info := range asinfo.All() {
+		if strings.Contains(strings.ToLower(info.Handle), query) ||
+			strings.Contains(strings.ToLower(info.Description), query) {
+			matches = append(matches, info)
+		}
+	}
+
+	if len(matches) <= limit {
+		t.Fatalf("%q matches %d entries; the test needs more than %d",
+			query, len(matches), limit)
+	}
+
+	slices.SortFunc(matches, func(a, b asinfo.Info) int {
+		return cmp.Compare(a.ASN, b.ASN)
+	})
+
+	results, total := asinfo.Search(query, limit)
+
+	if total != len(matches) {
+		t.Errorf("Search(%q, %d): total %d, want %d",
+			query, limit, total, len(matches))
+	}
+
+	if !slices.Equal(results, matches[:limit]) {
+		t.Errorf("Search(%q, %d) = %v, want %v",
+			query, limit, results, matches[:limit])
+	}
+}
+
 func TestDataIntegrity(t *testing.T) {
+	t.Parallel()
+
 	// Verify no duplicate ASNs
-	all := All()
+	all := asinfo.All()
 	seen := make(map[int]bool)
+
 	for _, info := range all {
 		if seen[info.ASN] {
 			t.Errorf("Duplicate ASN found: %d", info.ASN)
 		}
+
 		seen[info.ASN] = true
 	}
 
@@ -163,6 +286,7 @@ func TestDataIntegrity(t *testing.T) {
 		if info.Handle == "" && info.ASN != 0 {
 			t.Errorf("ASN %d has empty handle", info.ASN)
 		}
+
 		if info.Description == "" {
 			t.Errorf("ASN %d has empty description", info.ASN)
 		}
@@ -174,16 +298,20 @@ func BenchmarkGet(b *testing.B) {
 	asns := []int{1, 15169, 13335, 32934, 8075, 16509}
 
 	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
+
-		Get(asns[i%len(asns)])
+	for i := range b.N {
+		asinfo.Get(asns[i%len(asns)])
 	}
 }
 
 func BenchmarkSearch(b *testing.B) {
+	const limit = 50
+
 	queries := []string{"Google", "Amazon", "Microsoft", "University"}
 
 	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
+
-		Search(queries[i%len(queries)])
+	for i := range b.N {
+		asinfo.Search(queries[i%len(queries)], limit)
 	}
 }

pkg/asinfo/doc.go

@@ -2,7 +2,7 @@
 Package asinfo provides information about Autonomous Systems (AS).
 
 The package embeds a comprehensive database of AS numbers, handles, and descriptions
-sourced from https://github.com/ipverse/asn-info. The data is embedded at compile
+sourced from https://github.com/ipverse/as-metadata. The data is embedded at compile
 time and provides fast, offline lookups.
 
 Basic usage:
@@ -16,8 +16,10 @@ Basic usage:
 	// Get just the description
 	desc := asinfo.GetDescription(15169) // "Google LLC"
 
-	// Search for AS entries
+	// Search the handles and descriptions, ignoring case, for the first 10
-	results := asinfo.Search("University")
+	// matches in AS number order and the number of all matches
+	results, total := asinfo.Search("university", 10)
+	fmt.Printf("%d matches\n", total)
 	for _, as := range results {
 		fmt.Printf("AS%d: %s\n", as.ASN, as.Description)
 	}