Files
OwnCord/Server/api/constants.go
T
J3vbandClaude Fable 5 f5faf82a60 infra: observability, backups, guardrails, and deployment hardening (#1376)
* docs: add infrastructure roadmap plan

Records the verified recommendations from an infrastructure review in three
tracks: raising the single-instance ceiling, cheap seams for a possible
multi-instance future, and ops hygiene. Includes explicit anti-recommendations
and sequencing. Security-sensitive detail is intentionally excluded per
docs/security.md.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* feat(server): real health checks and saturation metrics

/api/v1/metrics now exposes signals that were already computed in memory but
never surfaced: reconnect replay tier hits, event-persister counters, SQLite
writer-pool wait stats, aggregate per-client backpressure counters (including
previously invisible low-priority drops), and permission-cache hit/miss.

/health now returns a real verdict: hub dispatch-loop liveness, a bounded
database ping, and a free-disk check, returning 503 with a subsystem reason
when degraded. Checks are cached so the unauthenticated endpoint cannot
amplify load. The hub's panic breaker now exits the process so a supervisor
can restart it, instead of leaving broadcast delivery silently dead while
clients still appear online.

OTel instruments that were declared but never recorded are now wired
(ws_active_connections, ws_broadcast_latency_seconds, ws_messages_total,
ws_events_dropped_total, voice gauges) or removed (db_query_duration_seconds).
Also corrects the docs/api.md description of broadcast_drops, which counts
hub-queue overflow, not client send-queue overflow.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* feat(server): implement scheduled backups, retention, and backup verification

The backup_schedule and backup_retention settings have existed in the admin
panel and API since the initial schema but were never read by any code. The
15-minute maintenance loop now enforces them: a scheduled backup is taken
when the newest backup on disk is older than the schedule interval (manual
backups reset the clock), and retention prunes backups older than the
configured days while always keeping the newest one.

Backups are now verified with PRAGMA integrity_check immediately after
VACUUM INTO (a failed backup is removed rather than listed as restorable)
and again before a restore may overwrite the live database. A failed VACUUM
INTO also cleans up its partial output file — but never a pre-existing one.

The backup directory is configurable via a new backup.dir key (default
data/backups) so operators can point backups at another disk or an off-host
mount, mirroring the SetDatabasePath plumb.

Restore-handler tests now use real SQLite fixtures (the integrity gate
correctly refuses text files) with the mid-copy failure injected through a
test-only copy hook. Also adds audited gosec suppressions to the Windows
disk-free syscall added in the previous commit, which the Windows lint leg
flagged.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* feat(server): capacity and failure-mode guardrails

- server.max_ws_connections: optional cap on concurrent WebSocket clients,
  checked before the upgrade with a 503 + Retry-After; rejections are counted
  and exposed as ws_conn_rejects in /api/v1/metrics.
- Single-process database lock: an OS-level advisory lock (flock / exclusive
  handle) beside the SQLite file makes a second server process fail fast with
  a clear message instead of silently fighting the first over process-local
  state. A bounded retry covers the self-update/restore restart handoff, and
  the lock mechanism failing (e.g. network filesystems) only warns.
- Disk-space awareness: boot-time warnings for the data and backup volumes,
  plus a disk_free_mb metrics field, via a small cross-platform diskutil
  package (already used by /health).
- Upload storage failures: storage.Save now marks server-side filesystem
  failures with a sentinel (storage.ErrIO); handlers return 507 for those
  instead of blaming the client with a 400, and the emoji route stops echoing
  raw storage errors (which embed absolute paths) into responses.
- Unknown config keys now warn at startup — a typo like admin_alowed_cidrs
  previously kept the default silently while the operator believed the
  setting changed. Never fatal: newer servers tolerate older configs.
- Admin settings honesty: the three stored-but-inert settings (server_icon,
  max_upload_bytes, voice_quality) are shown read-only with a note pointing
  at the real config.yaml keys, instead of pretending to apply.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* perf(db): write-path efficiency and capacity knobs

- channel_focus/mark_read now skip the read-state UPSERT when the stored row
  already matches (same last_message_id, no mentions) — refocus events fire
  at up to 10/s/user and every no-op write still occupied the single SQLite
  writer connection. The extra existence check runs on the reader pool, which
  doesn't serialize. Same shape as the session-touch throttle.
- DeleteExpiredSessions is now sargable: migration 031 normalizes legacy
  expiry formats to the RFC3339-Z layout the server writes and indexes
  expires_at, replacing the strftime full-table scan that ran on the writer
  every 15 minutes.
- Boot-time ANALYZE runs only when a migration actually applied; unchanged
  schemas get the cheap PRAGMA optimize instead (which also covers
  crash-restarts that never reached the shutdown optimize).
- The read/write SQL router gets a table-driven test with explicit expected
  values (INSERT ... RETURNING must hit the writer despite being :one).
- New knobs, all defaulting to current behavior: database.max_readers,
  security.auth_rate_limit_multiplier (for shared-NAT communities),
  event_persistence.replay_ring_size and replay_cold_limit.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* fix(server): shutdown lifecycle ordering

- The event pruner and maintenance loop are now joined (bounded) before the
  database closes: bgCtx cancellation used to run AFTER database.Close via
  LIFO defers, contradicting its own comment, and neither goroutine was ever
  waited on — a mid-tick scheduled backup or prune could still hold the
  writer while the pool tore down. StartEventPruner returns a done channel
  with the same join contract EventPersister.Stop already had.
- srv.Shutdown now runs before hub.GracefulStop, so in-flight HTTP handlers'
  broadcasts still reach a live hub and the event persister instead of
  vanishing from the replay/event store across a restart. Shutdown does not
  wait on hijacked WebSocket connections, so the swap adds no delay.
- GracefulStopContext threads the 30s shutdown budget into the hub: the 5s
  client-notice window (matching the countdown clients are shown) ends early
  when the budget expires, and is skipped entirely when nobody is connected —
  early-return startup paths and idle servers no longer sleep 5s for an
  audience of zero.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* build(deploy): systemd unit, compose hardening, boot-smoked releases, CI polish

- deploy/owncord.service: hardened systemd unit template with the two
  verified caveats encoded (install dir stays writable for self-update under
  ProtectSystem=strict; CAP_NET_BIND_SERVICE for ACME's :80), plus a
  'Linux (systemd)' deployment docs section — the Linux service story was
  previously 'Docker or nothing'.
- New 'Reverse Proxy Topology' docs section with a working nginx snippet and
  the correct signaling-vs-media distinction: /livekit/* is already proxied
  by the server, only WebRTC media ports must be directly reachable.
- docker-compose: log rotation, commented resource limits, and a healthcheck
  backed by a new 'chatserver healthcheck' subcommand (the distroless image
  has no shell) that probes /health without config side effects.
- release.yml: a concurrency group (queue, never cancel), and boot-smoke
  gates — the freshly built server binaries and the Docker image are cold
  booted and probed healthy BEFORE anything is signed or pushed. The release
  feed drives signed self-updates, so a binary that compiles but dies on
  boot previously would have shipped itself to every auto-updating instance.
- ci.yml: client-check/client-tests move to ubuntu with the reasoning
  recorded (no win32 code paths, LF enforced repo-wide); admin-e2e gets a
  written graduation criterion instead of an open-ended non-blocking status.
- docs: Tailscale guide notes the CGNAT range vs the default admin CIDRs;
  architecture overview records presence/voice state as the fifth
  single-instance blocker and the macOS client scope decision.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* perf(server): measured load tooling, narrowed invalidation, presence coalescing, storage and CIDR seams

- Fix scripts/k6/ws-load.js against the real wire protocol: envelope-wrapped
  frames, correct message types (typing_start, presence_update), the correct
  /api/v1/ws path, and thresholds that fail a run where nobody authenticated
  or went ready — the script had drifted to pre-envelope framing and reported
  100% green while every auth failed on the first frame. A new
  workflow_dispatch-only load-baseline workflow boots a real server, seeds
  users through the setup/invite APIs, runs the script, and uploads the k6
  summary plus a metrics snapshot for before/after comparison.
- Role-scoped channel-override changes now evict only the affected role's
  members from the permission cache (fail-safe: unreadable member list still
  flushes everything). InvalidateAll here repopulated every connected user —
  two reads each — synchronously inside the admin request via
  RefreshChannelVisibility, a stampede that scaled with total population
  rather than the role's size. Same pattern the per-user override endpoints
  already used.
- Connect/disconnect presence broadcasts now pass through a 300ms latest-wins
  coalescer (QueuePresence): each un-coalesced presence change is a sequenced
  global broadcast (an O(clients) fan-out under seqMu), so a reconnect storm
  fired O(users) of them from the connect critical path. A flap inside the
  window collapses to its final state; the wire format, seq ordering, and
  replay behaviour are unchanged, and the delivery path (BroadcastPresence)
  is untouched.
- Storage seam: api handlers now consume a FileStore interface (consumer-side,
  same pattern as service.Store) with Open returning a seekable storage.File —
  writing down the contract (range-request seeks included) an alternative
  backend would have to meet, without building one.
- The metrics surfaces and the LiveKit webhook/health endpoints get their own
  allowlist keys (metrics_allowed_cidrs, livekit_webhook_allowed_cidrs, both
  defaulting to admin_allowed_cidrs), so a central Prometheus scraper or an
  externally-hosted LiveKit no longer requires widening the admin panel's
  perimeter. Startup now also warns when admin_allowed_cidrs is customized
  while trusted_proxies is empty — behind a proxy or container network the
  check would otherwise compare the proxy's private address, not the client's.
- The container healthcheck probe now PINS the server's own certificate from
  disk (VerifyConnection, exact-match) instead of skipping TLS verification,
  addressing the CodeQL finding on the previous commit; WebPKI verification
  is used when no local cert exists (ACME).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* fix(server): address self-review findings on the hardening branch

Seven fixes from a high-effort review of the full branch diff:

- healthcheck CLI now works under tls.mode acme: it overrides ServerName
  with the configured domain for WebPKI verification instead of pinning a
  cert that doesn't exist (or is stale) in that mode. Previously an ACME
  deployment's container healthcheck failed forever.
- /health pings the READER pool (new db.PingRead): the writer ping queued
  behind a scheduled backup's VACUUM INTO and reported the server degraded
  for the whole backup — which an autoheal watchdog would turn into a
  nightly mid-backup restart.
- /health runs its cached checks under context.WithoutCancel so a probe
  that disconnects mid-request cannot poison the shared cache with a false
  degraded verdict for the next 5 seconds.
- The token CLI uses a new db.OpenShared that skips the single-process
  lock: minting a token against a running server is safe under WAL and was
  a documented workflow the lock had broken.
- The per-user TOTP failure cap is no longer scaled by
  security.auth_rate_limit_multiplier — that knob exists for per-IP limits;
  scaling the only cross-IP brute-force defence multiplied an attacker's
  distributed guess budget. Mirrors the unscaled per-user login threshold.
- A direct presence_update now drops the user's queued entry in the
  connect/disconnect coalescer, so a stale connect-time presence can no
  longer flush 300ms later over the user's fresher chosen status.
- The scheduled-backup filename collision loop breaks on any stat error
  and bounds its suffix probing, instead of spinning the maintenance
  goroutine forever on a persistent EACCES.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

* test(admin): real SQLite fixture for the merged Close-failure restore test

TestHandleRestoreBackup_RestartsWhenCloseFails arrived from main (#1375)
with a plain-text backup fixture; this branch's restore handler verifies
backups with integrity_check before touching the live database, so the text
fixture was (correctly) refused with 400 before the Close-failure branch
under test was reached. Use a real backup via BackupToSafe, matching the
other restore tests.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017RtDNHSYWwPKArL8MsRdbj

---------

Co-authored-by: Claude <noreply@anthropic.com>
2026-08-15 20:50:47 +02:00

265 lines
11 KiB
Go

package api
import (
"math"
"sync/atomic"
"time"
"github.com/owncord/server/config"
)
// ─── Rate limits ────────────────────────────────────────────────────────────
//
// Each constant defines either a request cap or a sliding-window duration used
// by the per-endpoint rate limiters.
// authRateScaleBits holds the security.auth_rate_limit_multiplier as float
// bits. It scales the per-IP auth request caps and failure thresholds for
// deployments where many users share one IP (office/school NAT) — the
// compiled-in constants below assume roughly one person per address. Atomic
// because tests construct multiple routers concurrently. Set via
// setAuthRateScale in NewRouter; reads happen at mount time and on the login
// failure-count path.
var authRateScaleBits atomic.Uint64
func init() { authRateScaleBits.Store(math.Float64bits(1.0)) }
// setAuthRateScale clamps and installs the auth rate multiplier. Zero or
// negative (unset config) means 1.0.
func setAuthRateScale(m float64) {
if m <= 0 {
m = 1.0
}
m = math.Min(math.Max(m, 0.1), 100)
authRateScaleBits.Store(math.Float64bits(m))
}
// scaledAuthLimit applies the auth rate multiplier to a compiled-in limit,
// never returning less than 1.
func scaledAuthLimit(n int) int {
scaled := int(math.Round(float64(n) * math.Float64frombits(authRateScaleBits.Load())))
if scaled < 1 {
return 1
}
return scaled
}
const (
// registerRateLimitPerMinute is the maximum registration attempts per IP per minute.
registerRateLimitPerMinute = 3
// loginRateLimitPerMinute is the maximum login attempts per IP per minute.
loginRateLimitPerMinute = 5
// verifyTOTPRateLimitPerMinute is the maximum TOTP verification attempts per IP per minute.
verifyTOTPRateLimitPerMinute = 10
// sensitiveEndpointRateLimitPerMinute is the rate limit applied to destructive
// or sensitive endpoints (account deletion, TOTP enable/confirm/disable).
sensitiveEndpointRateLimitPerMinute = 5
// searchRateLimitPerMinute is the maximum full-text search requests per IP per minute.
searchRateLimitPerMinute = 30
// livekitProxyRateLimitPerMinute is the maximum LiveKit proxy requests per IP per minute.
livekitProxyRateLimitPerMinute = 30
// clientUpdateRateLimitPerMinute is the maximum client-update checks per IP per minute.
clientUpdateRateLimitPerMinute = 30
// gifRateLimitPerMinute is the maximum GIF proxy requests per IP per minute.
// The picker debounces at 300ms, so a user typing continuously for a minute
// stays under this; it exists to bound abuse of the operator's Klipy quota.
gifRateLimitPerMinute = 30
// loginFailureThreshold is the number of failed login attempts (within
// loginFailureWindow) before the IP is locked out.
loginFailureThreshold = 9
// loginFailureWindow is the sliding window for counting login failures.
loginFailureWindow = 15 * time.Minute
// loginLockoutDuration is how long an IP is locked out after exceeding
// loginFailureThreshold.
loginLockoutDuration = 15 * time.Minute
// deleteAccountFailureThreshold is the number of wrong-password attempts
// before the per-user lockout kicks in.
deleteAccountFailureThreshold = 3
// deleteAccountFailureWindow is the sliding window for counting
// delete-account password failures.
deleteAccountFailureWindow = 15 * time.Minute
// deleteAccountLockoutDuration is how long the account-deletion endpoint
// is locked after exceeding deleteAccountFailureThreshold.
deleteAccountLockoutDuration = 15 * time.Minute
// totpFailureRateLimit is the maximum TOTP verification failures per user
// within totpFailureWindow before the user is rate-limited.
totpFailureRateLimit = 10
// totpFailureWindow is the sliding window for counting per-user TOTP failures.
totpFailureWindow = 15 * time.Minute
// partialAuthMaxFailures is the number of failed TOTP attempts on a single
// partial-auth challenge before it is revoked.
partialAuthMaxFailures = 5
// profilePasswordRateLimitPerMinute is the maximum password change attempts
// per IP per minute.
profilePasswordRateLimitPerMinute = 5
// profileUpdateRateLimitPerMinute is the maximum profile update attempts
// per user per minute.
profileUpdateRateLimitPerMinute = 10
// loginUserFailureThreshold is the number of failed login attempts for a
// specific username (regardless of source IP) before the account is locked.
loginUserFailureThreshold = 9
// loginUserFailureWindow is the sliding window for per-username login failures.
loginUserFailureWindow = 15 * time.Minute
// loginUserLockoutDuration is how long a username is locked after exceeding
// loginUserFailureThreshold.
loginUserLockoutDuration = 15 * time.Minute
// pwConfirmFailureThreshold is the number of wrong-password attempts on
// password-confirmation endpoints before per-user lockout kicks in.
pwConfirmFailureThreshold = 3
// pwConfirmFailureWindow is the sliding window for per-user password
// confirmation failures.
pwConfirmFailureWindow = 15 * time.Minute
// pwConfirmLockoutDuration is how long password-confirmation endpoints are
// locked after exceeding pwConfirmFailureThreshold.
pwConfirmLockoutDuration = 15 * time.Minute
// uploadRateLimitPerMinute is the maximum file uploads per user per minute.
uploadRateLimitPerMinute = 10
// emojiUploadRateLimitPerMinute is the maximum custom-emoji uploads per
// MANAGE_SERVER holder per minute. Lower than the attachment limit: every
// accepted upload fans an emoji_update out to every connected session.
emojiUploadRateLimitPerMinute = 10
)
// ─── Timeouts & TTLs ────────────────────────────────────────────────────────
const (
// partialAuthStoreTTL is the lifetime of a partial-auth (2FA) challenge token.
partialAuthStoreTTL = 10 * time.Minute
// pendingTOTPStoreTTL is the lifetime of a pending TOTP enrollment secret.
pendingTOTPStoreTTL = 10 * time.Minute
// rateLimiterCleanupInterval is how often stale rate-limiter entries are reaped.
rateLimiterCleanupInterval = 5 * time.Minute
// rateLimiterCleanupMaxWindow is the maximum window considered when pruning
// stale rate-limiter entries. It must cover the LARGEST window any caller
// passes to Allow: slow mode (service/message_crud.go) uses windows up to
// admin's maxSlowModeSeconds (21600 s = 6 h), and a shorter horizon makes
// the reaper silently reset long slow modes after ~15 minutes.
rateLimiterCleanupMaxWindow = 6 * time.Hour
// hstsMaxAgeSeconds is the max-age value for the Strict-Transport-Security header.
hstsMaxAgeSeconds = 31536000
)
// bodyCapExemptPrefixes are the route prefixes excluded from the global 1 MiB
// body cap because they enforce their own, larger envelope at the route or
// handler level. A route with a documented cap above 1 MiB that is missing
// here is unreachable at its own limit: MaxBytesReader wrappers merely
// delegate reads, so the innermost (global) limit errors first.
var bodyCapExemptPrefixes = []string{
"/api/v1/uploads",
// 16 MiB plugin envelope enforced by the handler's own MaxBytesReader.
"/api/v1/admin/plugins/install",
// 2 MiB avatar envelope: route-scoped MaxBodySize(avatarMaxBodySize)
// plus the handler's re-wrap enforce it.
"/api/v1/users/me/avatar",
}
// ─── Size limits ────────────────────────────────────────────────────────────
const (
// defaultMaxBodySize is the default request body size limit (1 MiB).
defaultMaxBodySize = config.MaxMessageBytes
// uploadMaxBodySize is the request body size limit for file uploads (100 MiB).
uploadMaxBodySize = 100 << 20
// multipartMemoryLimit is the in-memory limit for multipart form parsing;
// data beyond this is spilled to disk.
multipartMemoryLimit = 10 << 20
// maxUploadFilenameLength is the maximum length of an upload filename
// (filesystem-safe limit).
maxUploadFilenameLength = 255
// maxAvatarURLLen is the maximum length of a user avatar URL.
maxAvatarURLLen = 512
// maxEmojiFileBytes is the largest custom-emoji image accepted (512 KiB).
// An emoji renders at 22px inline and 48px jumbo, so anything approaching
// this is already far more data than the pixels can use.
maxEmojiFileBytes = 512 << 10
// maxEmojiDimension caps an emoji's width and height in pixels. Discord
// normalizes to 128px; matching it means an emoji uploaded for OwnCord
// looks the same as the one it was copied from, jumbo included.
maxEmojiDimension = 128
// emojiMaxBodySize bounds the whole multipart request. The image cap plus
// the form's own framing (boundaries, headers, the shortcode field) —
// generous enough that a legitimate 512 KiB upload never trips it.
emojiMaxBodySize = 1 << 20
// emojiMultipartMemoryLimit is the in-memory limit for parsing an emoji
// upload. Above maxEmojiFileBytes, so a valid emoji never spills to disk.
emojiMultipartMemoryLimit = 1 << 20
// maxAvatarFileBytes is the largest avatar image accepted (1 MiB). An
// avatar renders at 40px in a message row and 64px in the profile popup,
// so this is already generous; the client downscales before uploading and
// the cap is what stops it being used as free file hosting.
maxAvatarFileBytes = 1 << 20
// maxAvatarDimension caps an avatar's stored width and height. Bigger than
// any surface renders it, so a retina display still has pixels to spare
// while a 6000px camera JPEG is refused rather than shipped to every
// client that sees the user post.
maxAvatarDimension = 1024
// avatarMaxBodySize bounds the whole multipart avatar request: the image
// cap plus room for the form's boundaries and headers.
avatarMaxBodySize = 2 << 20
// avatarMultipartMemoryLimit is the in-memory limit for parsing an avatar
// upload. Above maxAvatarFileBytes, so a valid avatar never spills to disk.
avatarMultipartMemoryLimit = 2 << 20
// avatarUploadRateLimitPerMinute is the maximum avatar uploads per user per
// minute. Lower than the attachment limit: every accepted upload fans a
// user_update out to every connected session and orphans the previous file.
avatarUploadRateLimitPerMinute = 5
// maxRequestIDLen bounds a client-supplied X-Request-Id. chi's
// middleware.RequestID adopts that header verbatim, and the value then
// reaches every log record for the request (logctx, requestLogger,
// recoverer) and the echoed response header — while the admin ring buffer
// retains 2000 records, so an unbounded id becomes long-lived heap.
// 128 bytes fits every common correlation-id format (UUID, 32-hex,
// W3C traceparent, chi's own "host/prefix-000001").
maxRequestIDLen = 128
// maxLoggedPathLen bounds the request path attached to a log record. The
// URL is client-controlled and net/http accepts one up to MaxHeaderBytes
// (~1 MiB), so an unbounded path fills the ring buffer the same way an
// unbounded request id does. Well past the longest real route.
maxLoggedPathLen = 256
)