fix(backend): report ingest correctness — propagate storage failure, 413 on oversize, global body cap (closes #23)

A buffer failure on POST /api/v1/reports now returns 500 instead of a
false {"status":"ok"}, so clients can retry. Decode errors are split:
an over-limit body returns 413 (via errors.As on *http.MaxBytesError),
malformed JSON stays 400. A new MaxBodyBytes middleware caps every
route (1 MiB default; rejects an oversized Content-Length up-front and
caps the read otherwise), replacing the per-route reader so the health
check and future routes are bounded too. The raw attacker-controlled
geo blob is no longer logged — only its byte length — and client_id and
timestamp are length-bounded before logging. A decodeJSON handler helper
is added per the HTTP server conventions. Panic recovery is now a local
middleware that routes the stack through slog as structured JSON rather
than chi's plain-text stderr. Error bodies still leak nothing internal.

Chosen failure code for a storage failure: 500, since a full buffer or
write error is server-side and retryable, not the client's fault.

Model: opus-4-8

backend/internal/handlers/export_test.go

@@ -0,0 +1,10 @@
+package handlers
+
+import "log/slog"
+
+// NewForTest builds a Handlers around a report sink and logger,
+// bypassing the fx graph so handler behaviour (including the
+// storage failure path) is exercisable in unit tests.
+func NewForTest(buf reportAppender, log *slog.Logger) *Handlers {
+	return &Handlers{buf: buf, log: log}
+}

backend/internal/handlers/handlers.go

@@ -18,6 +18,13 @@ import (
 
 const jsonContentType = "application/json; charset=utf-8"
 
+// reportAppender is the subset of the report buffer the handlers
+// depend on. Defining it here keeps the storage failure path
+// exercisable with a stub in tests.
+type reportAppender interface {
+	Append(v any) error
+}
+
 // Params defines the dependencies for Handlers.
 type Params struct {
 	fx.In
@@ -30,7 +37,7 @@ type Params struct {
 
 // Handlers provides HTTP handler factories for all endpoints.
 type Handlers struct {
-	buf    *reportbuf.Buffer
+	buf    reportAppender
 	hc     *healthcheck.Healthcheck
 	log    *slog.Logger
 	params *Params
@@ -72,3 +79,15 @@ func (s *Handlers) respondJSON(
 		}
 	}
 }
+
+// decodeJSON decodes the request body into v. The body is
+// expected to already be bounded by the body-size middleware, so
+// a caller can distinguish an over-limit body from malformed
+// JSON by testing the returned error for *http.MaxBytesError.
+func (s *Handlers) decodeJSON(
+	_ http.ResponseWriter,
+	r *http.Request,
+	v any,
+) error {
+	return json.NewDecoder(r.Body).Decode(v)
+}

backend/internal/handlers/report.go

@@ -2,10 +2,14 @@ package handlers
 
 import (
 	"encoding/json"
+	"errors"
 	"net/http"
 )
 
-const maxReportBodyBytes = 1 << 20 // 1 MiB
+// maxLoggedFieldBytes bounds untrusted string fields before they
+// are logged, so a caller cannot inflate log volume with an
+// oversized value.
+const maxLoggedFieldBytes = 128
 
 type reportSample struct {
 	T       int64   `json:"t"`
@@ -35,48 +39,76 @@ func (s *Handlers) HandleReport() http.HandlerFunc {
 	}
 
 	return func(w http.ResponseWriter, r *http.Request) {
-		r.Body = http.MaxBytesReader(
-			w, r.Body, maxReportBodyBytes,
-		)
-
 		var rpt report
 
-		err := json.NewDecoder(r.Body).Decode(&rpt)
+		err := s.decodeJSON(w, r, &rpt)
 		if err != nil {
-			s.log.Error("failed to decode report",
-				"error", err,
-			)
 			s.respondJSON(w, r,
 				&response{Status: "error"},
-				http.StatusBadRequest,
+				s.decodeErrorStatus(err),
 			)
 
 			return
 		}
 
-		totalSamples := 0
-		for _, h := range rpt.Hosts {
-			totalSamples += len(h.History)
-		}
+		s.logReportReceived(rpt)
 
-		s.log.Info("report received",
-			"client_id", rpt.ClientID,
-			"timestamp", rpt.Timestamp,
-			"host_count", len(rpt.Hosts),
-			"total_samples", totalSamples,
-			"geo", string(rpt.Geo),
-		)
-
-		bufErr := s.buf.Append(rpt)
-		if bufErr != nil {
-			s.log.Error("failed to buffer report",
-				"error", bufErr,
+		err = s.buf.Append(rpt)
+		if err != nil {
+			s.log.Error("failed to buffer report", "error", err)
+			s.respondJSON(w, r,
+				&response{Status: "error"},
+				http.StatusInternalServerError,
 			)
+
+			return
 		}
 
-		s.respondJSON(w, r,
-			&response{Status: "ok"},
-			http.StatusOK,
-		)
+		s.respondJSON(w, r, &response{Status: "ok"}, http.StatusOK)
 	}
 }
+
+// decodeErrorStatus logs a report decode failure and returns the
+// status to send: 413 when the body exceeded the size limit,
+// otherwise 400 for malformed JSON.
+func (s *Handlers) decodeErrorStatus(err error) int {
+	var tooLarge *http.MaxBytesError
+	if errors.As(err, &tooLarge) {
+		s.log.Warn("report body too large", "limit_bytes", tooLarge.Limit)
+
+		return http.StatusRequestEntityTooLarge
+	}
+
+	s.log.Error("failed to decode report", "error", err)
+
+	return http.StatusBadRequest
+}
+
+// logReportReceived logs an accepted report. Untrusted fields are
+// bounded (client_id, timestamp) or reduced to a length
+// (geo_bytes) so the raw attacker-controlled body never reaches
+// the log.
+func (s *Handlers) logReportReceived(rpt report) {
+	totalSamples := 0
+	for _, h := range rpt.Hosts {
+		totalSamples += len(h.History)
+	}
+
+	s.log.Info("report received",
+		"client_id", boundedForLog(rpt.ClientID),
+		"timestamp", boundedForLog(rpt.Timestamp),
+		"host_count", len(rpt.Hosts),
+		"total_samples", totalSamples,
+		"geo_bytes", len(rpt.Geo),
+	)
+}
+
+// boundedForLog truncates an untrusted string to a fixed byte
+// bound so an attacker-controlled field cannot dominate the log.
+func boundedForLog(s string) string {
+	if len(s) > maxLoggedFieldBytes {
+		return s[:maxLoggedFieldBytes]
+	}
+
+	return s
+}

backend/internal/handlers/report_test.go

@@ -0,0 +1,141 @@
+package handlers_test
+
+import (
+	"bytes"
+	"encoding/json"
+	"errors"
+	"io"
+	"log/slog"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+
+	"sneak.berlin/go/netwatch/internal/handlers"
+)
+
+var errStorageFailed = errors.New("storage failed")
+
+// stubAppender drives the storage success/failure path without a
+// real buffer or disk.
+type stubAppender struct {
+	err error
+}
+
+func (s stubAppender) Append(any) error { return s.err }
+
+func newTestHandlers(buf stubAppender, out io.Writer) *handlers.Handlers {
+	return handlers.NewForTest(buf, slog.New(slog.NewJSONHandler(out, nil)))
+}
+
+func decodeStatus(t *testing.T, body []byte) string {
+	t.Helper()
+
+	var resp struct {
+		Status string `json:"status"`
+	}
+
+	err := json.Unmarshal(body, &resp)
+	if err != nil {
+		t.Fatalf("response body not JSON: %v (%q)", err, body)
+	}
+
+	return resp.Status
+}
+
+func TestHandleReportStorageFailureIsNon2xx(t *testing.T) {
+	t.Parallel()
+
+	h := newTestHandlers(stubAppender{err: errStorageFailed}, io.Discard)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(
+		http.MethodPost, "/api/v1/reports",
+		strings.NewReader(`{"clientId":"c1","hosts":[]}`),
+	)
+
+	h.HandleReport().ServeHTTP(rec, req)
+
+	if rec.Code < 500 {
+		t.Fatalf("storage failure status = %d, want a 5xx", rec.Code)
+	}
+
+	if got := decodeStatus(t, rec.Body.Bytes()); got != "error" {
+		t.Fatalf("status field = %q, want %q", got, "error")
+	}
+}
+
+func TestHandleReportMalformedJSONIs400(t *testing.T) {
+	t.Parallel()
+
+	h := newTestHandlers(stubAppender{}, io.Discard)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(
+		http.MethodPost, "/api/v1/reports",
+		strings.NewReader(`{not json`),
+	)
+
+	h.HandleReport().ServeHTTP(rec, req)
+
+	if rec.Code != http.StatusBadRequest {
+		t.Fatalf("malformed status = %d, want %d",
+			rec.Code, http.StatusBadRequest)
+	}
+}
+
+func TestHandleReportOversizeIs413(t *testing.T) {
+	t.Parallel()
+
+	const limit = 32
+
+	h := newTestHandlers(stubAppender{}, io.Discard)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(
+		http.MethodPost, "/api/v1/reports",
+		strings.NewReader(`{"clientId":"`+strings.Repeat("x", 200)+`"}`),
+	)
+	// Emulate the body-size middleware capping the body so the
+	// handler observes a *http.MaxBytesError while decoding.
+	req.Body = http.MaxBytesReader(rec, req.Body, limit)
+
+	h.HandleReport().ServeHTTP(rec, req)
+
+	if rec.Code != http.StatusRequestEntityTooLarge {
+		t.Fatalf("oversize status = %d, want %d",
+			rec.Code, http.StatusRequestEntityTooLarge)
+	}
+}
+
+func TestHandleReportDoesNotLogRawGeo(t *testing.T) {
+	t.Parallel()
+
+	const sentinel = "SENSITIVE-GEO-BLOB"
+
+	var logbuf bytes.Buffer
+
+	h := newTestHandlers(stubAppender{}, &logbuf)
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequest(
+		http.MethodPost, "/api/v1/reports",
+		strings.NewReader(
+			`{"clientId":"c1","geo":{"raw":"`+sentinel+`"},"hosts":[]}`,
+		),
+	)
+
+	h.HandleReport().ServeHTTP(rec, req)
+
+	if rec.Code != http.StatusOK {
+		t.Fatalf("status = %d, want %d", rec.Code, http.StatusOK)
+	}
+
+	if strings.Contains(logbuf.String(), sentinel) {
+		t.Fatal("raw geo bytes were written to the log")
+	}
+
+	if !strings.Contains(logbuf.String(), "geo_bytes") {
+		t.Fatal("expected a bounded geo_bytes field in the log")
+	}
+}