Files
silo-server/internal/logredact/redact.go
203a18ae83 feat(observability): OpenTelemetry logs+traces with secret redaction and slog standardization (#290)
* feat(observability): OpenTelemetry logs+traces with secret redaction

Part of #265. Adds opt-in OpenTelemetry (logs + traces) alongside the existing
stderr + opslog pipeline, plus secret redaction on all sinks. Default-off: with
no OTEL_* / SILO_OTEL_ENABLED config, behavior is unchanged.

Bootstrap (internal/telemetry):
- Setup() builds one shared resource, a TracerProvider (parent-based trace-id
  ratio sampler), a LoggerProvider, and the W3C TraceContext+Baggage propagator
  from env. It installs NO MeterProvider — metrics stay on Prometheus, and the
  built-in no-op global MeterProvider keeps the trace instrumentation libs from
  double-emitting. Shutdown is deferred with a flush timeout.
- Logs are bridged via otelslog fan-out (slog.MultiHandler), level-gated by the
  shared LevelVar and best-effort so a failing collector can't break the console
  or DB branches. stderr + opslog stay untouched.

Secret redaction (internal/logredact):
- A slog.Handler masks secret-keyed attributes (password, token, api_key,
  authorization, cookie, ...) — including .With-bound attrs, nested groups,
  secret-keyed group subtrees, and values behind a LogValuer — on the console
  and OTLP sinks, with a no-op fast path when a record has no secret keys.
  opslog.shouldRedact delegates to logredact.SecretKey so all sinks share one
  marker list.

Rotation is infra-managed (no custom file sink): container runtime for stderr,
collector/backend for OTLP, opslog partition-pruning for the DB. Documented in
docs/architecture/observability.md.

Verification: go build ./..., go vet, gofmt -l — clean; go test
./internal/telemetry/ ./internal/logredact/ -race pass.

AI-use disclosure: implemented with AI assistance (Claude Code), including
adversarial reviews that hardened the bootstrap and fixed two redaction leak
paths; reviewed by the author.

* refactor(observability): slog context+component sweep, sloglint gate (phase 3)

Part of #265. Builds on the OTel bootstrap + redaction commit.

Standardizes every log call site onto the context-carrying slog variants so
records correlate with the active OpenTelemetry trace, and locks the standard
in with a machine gate so future code (human- or AI-authored) can't drift back.

- Call-site sweep: converted the remaining slog.<Level>(...) calls to the
  slog.<Level>Context(ctx, ...) form wherever a context.Context is in scope
  (background/init calls with no ctx are left as-is), across 183 files. Applied
  via a type-aware AST codemod. Log levels and message strings are preserved
  verbatim; a component attr (canonical per-package name) is added to direct
  package-level slog calls. Bound-logger calls keep their existing .With
  bindings. The main.go and telemetry package conversions rode with their file
  in the previous commit to keep each file within a single commit.
- Enforcement (.golangci.yml): enable sloglint with context=scope, static-msg,
  key-naming-case=snake, no-mixed-args. After the sweep all four report zero
  violations repo-wide (tests included), so make lint / CI now blocks any
  regression to the non-context form. The gate ships with the sweep because it
  cannot be green until the legacy sites are converted.

Metrics remain on Prometheus; no behavior change to /metrics or Grafana.

Verification: go build ./..., go vet ./..., gofmt -l — clean; sloglint (all 4
rules) 0 violations repo-wide; log levels verified unchanged.

AI-use disclosure: implemented with AI assistance (Claude Code), including the
codemod; reviewed by the author.

* fix(observability): honor per-signal OTLP protocol and secret WithGroup names

Two Codex review findings on PR #290:

- telemetry: OTEL_EXPORTER_OTLP_{TRACES,LOGS}_PROTOCOL now override the
  generic OTEL_EXPORTER_OTLP_PROTOCOL per signal, so mixed collector
  setups (e.g. HTTP logs + gRPC traces) build the right exporter.
- logredact: entering a group whose name is secret-bearing (e.g.
  WithGroup("authorization")) now masks every leaf in that subtree,
  matching how slog.Group("authorization", ...) is masked as a whole.

* fix(observability): address review feedback on telemetry bootstrap

- Telemetry setup failure no longer kills boot: Setup returns usable
  no-op providers alongside the error and main logs and continues with
  telemetry disabled, honoring the best-effort contract.
- Honor OTEL_TRACES_SAMPLER (always_on/off, traceidratio, parentbased_*
  variants); unsupported values fall back to parentbased_traceidratio.
- Attach node identity as semconv service.instance.id instead of the
  non-semconv node.name.
- Rename opslog retention-scope log attrs to target_component/target_level
  so they no longer collide with the canonical component routing key, and
  tag those lines with component=opslog.
- Fix stale levelGated comment casing; use WarnContext in the telemetry
  shutdown defer; document the LogValuer double-resolve on the redaction
  slow path.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Quick <31828688+Quick104@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-09 08:53:52 -04:00

161 lines
5.6 KiB
Go

// Package logredact provides secret redaction for slog output sinks (console,
// file, OTLP). The opslog database path does its own redaction when flattening
// records into rows; this package covers the remaining sinks so secrets do not
// leak to stderr, a mounted log file, or an OTLP collector.
//
// Redaction is key-based: an attribute whose key looks secret-bearing (token,
// password, api_key, ...) has its value replaced with the placeholder. Values
// are not scanned, matching the opslog DB-path behavior — a secret embedded in
// a free-text message or a non-secret-keyed value is not caught.
package logredact
import (
"context"
"log/slog"
"strings"
)
// Placeholder replaces the value of a redacted attribute.
const Placeholder = "[REDACTED]"
// secretMarkers are substrings that, when present in a lower-cased attribute
// key, mark the value as secret-bearing. Kept in sync with the opslog DB path
// by being the single shared source (opslog.shouldRedact delegates here).
var secretMarkers = []string{
"password", "secret", "token", "api_key", "apikey", "authorization", "cookie",
}
// SecretKey reports whether an attribute key names a secret-bearing value.
func SecretKey(key string) bool {
key = strings.ToLower(key)
for _, marker := range secretMarkers {
if strings.Contains(key, marker) {
return true
}
}
return false
}
// Handler wraps an slog.Handler, redacting secret-bearing attributes (including
// those bound via WithAttrs and nested groups) before they reach the inner
// handler. It preserves level gating and best-effort semantics of whatever it
// wraps by delegating Enabled/Handle to the inner handler.
type Handler struct {
inner slog.Handler
// redactAll is set once the logger enters a group whose name is
// secret-bearing (e.g. WithGroup("authorization")). Inside such a subtree
// every leaf is masked regardless of its own key, matching how a
// slog.Group("authorization", ...) value is masked as a whole. It stays set
// for all descendant groups.
redactAll bool
}
// New returns a Handler wrapping inner.
func New(inner slog.Handler) *Handler {
return &Handler{inner: inner}
}
// Enabled delegates to the inner handler.
func (h *Handler) Enabled(ctx context.Context, level slog.Level) bool {
return h.inner.Enabled(ctx, level)
}
// Handle redacts the record's attributes and forwards to the inner handler.
// When the record carries no secret-bearing keys, the original record is passed
// through unchanged to avoid rebuilding it on the hot path.
func (h *Handler) Handle(ctx context.Context, r slog.Record) error {
if !h.redactAll && !recordNeedsRedaction(r) {
return h.inner.Handle(ctx, r)
}
nr := slog.NewRecord(r.Time, r.Level, r.Message, r.PC)
r.Attrs(func(a slog.Attr) bool {
nr.AddAttrs(redactAttr(a, h.redactAll))
return true
})
return h.inner.Handle(ctx, nr)
}
// WithAttrs redacts the bound attributes so secrets attached via a logger's
// With(...) are also masked.
func (h *Handler) WithAttrs(attrs []slog.Attr) slog.Handler {
redacted := make([]slog.Attr, len(attrs))
for i, a := range attrs {
redacted[i] = redactAttr(a, h.redactAll)
}
return &Handler{inner: h.inner.WithAttrs(redacted), redactAll: h.redactAll}
}
// WithGroup delegates to the inner handler; grouped attributes are still
// redacted by leaf key at Handle/WithAttrs time. A group whose name is itself
// secret-bearing masks every leaf within the subtree, so nothing logged under
// it (e.g. WithGroup("authorization").Info(..., "value", token)) leaks.
func (h *Handler) WithGroup(name string) slog.Handler {
return &Handler{
inner: h.inner.WithGroup(name),
redactAll: h.redactAll || SecretKey(name),
}
}
// recordNeedsRedaction reports whether any record-level attribute (or nested
// group attribute) has a secret-bearing key.
func recordNeedsRedaction(r slog.Record) bool {
found := false
r.Attrs(func(a slog.Attr) bool {
if attrHasSecret(a) {
found = true
return false
}
return true
})
return found
}
func attrHasSecret(a slog.Attr) bool {
// Resolve LogValuer values so a secret hidden behind one (e.g. a type whose
// LogValue() returns a token) is inspected, not passed through opaque.
a.Value = a.Value.Resolve()
// A secret-bearing key masks the whole attribute — including a group whose
// own key is secret (its members are never reached below).
if SecretKey(a.Key) {
return true
}
if a.Value.Kind() == slog.KindGroup {
for _, ga := range a.Value.Group() {
if attrHasSecret(ga) {
return true
}
}
}
return false
}
// redactAttr masks a's value when its key is secret-bearing. When force is set
// (the logger is inside a secret-named group), every leaf is masked regardless
// of its own key, while group structure is preserved so the subtree stays
// well-formed.
func redactAttr(a slog.Attr, force bool) slog.Attr {
// On the Handle slow path this re-invokes any LogValuer that attrHasSecret
// already resolved (on its own copy). slog requires LogValue to be cheap and
// side-effect free, so the double call is acceptable and is what keeps the
// no-secrets fast path allocation-free.
a.Value = a.Value.Resolve()
if a.Value.Kind() == slog.KindGroup {
// A group whose own key is secret collapses to a single placeholder,
// matching the leaf case; otherwise recurse, propagating force.
if SecretKey(a.Key) {
return slog.String(a.Key, Placeholder)
}
group := a.Value.Group()
out := make([]slog.Attr, len(group))
for i, ga := range group {
out[i] = redactAttr(ga, force)
}
return slog.Attr{Key: a.Key, Value: slog.GroupValue(out...)}
}
// Leaf: mask when forced or when the key itself is secret-bearing.
if force || SecretKey(a.Key) {
return slog.String(a.Key, Placeholder)
}
return a
}