An operator running `sqlite3 <db> .dump` against their own per-webhook
database wedged it: inbound webhooks rejected with HTTP 500, delivered
webhooks stranded at `pending`, and every one of them POSTed a second
time on the next restart while the event log recorded a single attempt.
Durability. Every SQLite file — main, per-webhook, and archive — now
opens through one path, `internal/database/sqlite_open.go`, in WAL
journal mode with a 10-second busy timeout, `BEGIN IMMEDIATE`
transactions, and a bounded connection pool. WAL is what stops a reader
blocking writers at all. `_txlock=immediate` is what stops a `COMMIT`
failing while its transaction stays open on a pooled connection, which
is how four `database is locked` errors became 593 `cannot start a
transaction within a transaction`. `cache=shared` is gone, because
under it an in-process conflict is SQLITE_LOCKED, which the busy
handler does not retry. The busy timeout is applied before
journal_mode: the driver runs DSN pragmas in order on every new
connection, and `PRAGMA journal_mode` takes a lock, so the reverse
order leaves the one pragma that can block uncovered by the handler
meant to cover it.
Eligibility. `internal/delivery/inflight.go` holds the set of
deliveries the engine owns — taken when a task is queued, when a
target schedules a retry, and by every recovery path before it
re-dispatches; dropped when the worker that ran the task returns.
Recovery and both sweep arms re-dispatch only what the set does not
hold. Nothing decides that from a row's age: a delivery waiting in a
10000-deep channel is arbitrarily old and perfectly healthy, and
reasoning from age re-sends it. `takeForRedispatch` is the single gate
every re-dispatch goes through — ownership first, then a conditional
update confirming the row is still in the status the batch read.
Bookkeeping. `recordResult` and `updateDeliveryStatus` return their
errors instead of logging and dropping them, and a caller whose
bookkeeping write failed writes nothing at all: the delivery keeps
whichever non-terminal status it already held, and the sweeps recover
it. Every recovery path — pending and retrying alike — first settles
any delivery that already holds a successful `DeliveryResult` rather
than sending it again. Recovery continues each delivery's own attempt
numbering instead of restarting at 1. The sweep gains a
`pending`-with-age-bound arm, so a stranded delivery no longer waits
for a restart.
Docs. WAL produces `-wal`/`-shm` sidecars, so the backup and restore
procedures in README.md are corrected against measurement: both
documented procedures were re-run against a live instance, a `-wal`
left by a crash carries data the `.db` alone does not, and an archive
file normally holds its rows in a `-wal` rather than in the `.db`.
The Status section claimed next held the completed 1.0.0 milestone with
every issue closed. Four are open (#176, #178, #186, #187), so a merge
of next to main would have shipped that claim to main.
Next Step still named #115 and #125 as open owner decisions; both
landed. It now names the real open items, #150 and #112, and the forced
merge order for the remaining four.
Completed Steps was seventeen units behind, back to 2026-08-12.
DBURL → DATA_DIR consolidation:
- Remove DBURL env var entirely; main DB now lives at {DATA_DIR}/webhooker.db
- database.go constructs DB path from config.DataDir, ensures dir exists
- Update DATA_DIR prod default from /data/events to /data
- Update all tests to use DataDir instead of DBURL
- Update Dockerfile: /data (not /data/events) for all SQLite databases
- Update README configuration table, Docker examples, architecture docs
Dead code removal:
- Remove unused IndexResponse struct (handlers/index.go)
- Remove unused TemplateData struct (handlers/handlers.go)
Stale comment cleanup:
- Remove TODO in server.go (DB cleanup handled by fx lifecycle)
- Fix nolint:golint → nolint:revive on ServerParams for consistency
- Clean up verbose middleware/routing comments in routes.go
- Fix TODO fan-out description (worker pool, not goroutine-per-target)
.gitignore fixes:
- Add data/ directory to gitignore
- Remove stale config.yaml entry (env-only config since rework)
The webhook handler now builds DeliveryTask structs carrying all target
config and event data inline (for bodies ≤16KB) and sends them through
the delivery channel. In the happy path, the engine delivers without
reading from any database — it only writes to record delivery results.
For large bodies (≥16KB), Body is nil and the engine fetches it from the
per-webhook database on demand. Retry timers also carry the full
DeliveryTask, so retries avoid unnecessary DB reads.
The database is used for crash recovery only: on startup the engine scans
for interrupted pending/retrying deliveries and re-queues them.
Implements owner feedback from issue #15:
> the message in the <=16KB case should have everything it needs to do
> its delivery. it shouldn't touch the db until it has a success or
> failure to record.
Split data storage into main application DB (config only) and
per-webhook event databases (one SQLite file per webhook).
Architecture changes:
- New WebhookDBManager component manages per-webhook DB lifecycle
(create, open, cache, delete) with lazy connection pooling via sync.Map
- Main DB (DBURL) stores only config: Users, Webhooks, Entrypoints,
Targets, APIKeys
- Per-webhook DBs (DATA_DIR) store Events, Deliveries, DeliveryResults
in files named events-{webhook_uuid}.db
- New DATA_DIR env var (default: ./data dev, /data/events prod)
Behavioral changes:
- Webhook creation creates per-webhook DB file
- Webhook deletion hard-deletes per-webhook DB file (config soft-deleted)
- Event ingestion writes to per-webhook DB, not main DB
- Delivery engine polls all per-webhook DBs for pending deliveries
- Database target type marks delivery as immediately successful (events
are already in the dedicated per-webhook DB)
- Event log UI reads from per-webhook DBs with targets from main DB
- Existing webhooks without DB files get them created lazily
Removed:
- ArchivedEvent model (was a half-measure, replaced by per-webhook DBs)
- Event/Delivery/DeliveryResult removed from main DB migrations
Added:
- Comprehensive tests for WebhookDBManager (create, delete, lazy
creation, delivery workflow, multiple webhooks, close all)
- Dockerfile creates /data/events directory
README updates:
- Per-webhook event databases documented as implemented (was Phase 2)
- DATA_DIR added to configuration table
- Docker instructions updated with data volume mount
- Data model diagram updated
- TODO updated (database separation moved to completed)
Closes#15
clawbot