c0ecf2fa3bebd46bdfd0169efd73b89653d1dfed
2
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 bjornbergenheimandClaude Opus 5
|
43a71aec3c |
fix(supabase): stop server clients leaking a 30s refresh ticker per request (#1612)
* fix(supabase): stop server clients leaking a 30s refresh ticker per request
`autoRefreshToken` defaults to true in supabase-js, and off-browser
@supabase/auth-js starts the refresh ticker unconditionally:
// in non-browser environments the refresh token ticker runs always
this.startAutoRefresh()
That is a setInterval firing every 30 s. It calls unref(), so the process
still exits, tests pass, and Vercel never notices because the process is
torn down long before the tickers accumulate. But unref() does not make a
timer collectable: it stays registered in the event loop and remains a GC
root for its callback, which closes over the GoTrueClient, the
SupabaseClient, and the whole request scope around it.
A long-running self-hosted instance therefore leaks one timer plus one
entire request graph (socket, IncomingMessage, ServerResponse, headers,
route context: ~100 kB) per client constructed. One died of "JavaScript
heap out of memory" after 42 h, the last 24 of them completely idle. The
heap snapshot showed 445 retained request graphs and ~1050 Timeouts in
the 30 000 ms bucket, retained via `autoRefreshTicker`, and the rate
matched the traffic exactly: the Docker healthcheck polls /api/health
every 30 s and the webhook dispatch cron runs every minute, so
3 clients/min x 148 min = 444.
- new lib/supabase/service-client.ts: createServiceRoleClient() applies
SERVER_AUTH_OPTIONS, spread LAST so a caller passing its own auth block
cannot re-enable the ticker
- 22 call sites migrated; only booking-templates/sync/cron had ever
passed the options itself
- guard 9 in no-new-antipatterns.mjs fails CI on any new value import of
supabase-js's createClient outside the wrapper; type-only imports are
fine. Verified to fail on a deliberate regression and pass once fixed
- browser clients untouched: a signed-in tab genuinely needs the refresh,
and lib/supabase/client.ts is built on createBrowserClient anyway
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
* fix(checks): catch namespace imports in the leaky-supabase-client guard
The guard only matched named imports, so
import * as sb from '@supabase/supabase-js'
sb.createClient(url, key)
reached createClient through member access without ever naming it, and
passed. Verified against the real script before and after: the shape is
flagged now, and `import type * as sb` still passes.
Namespace value imports are treated as leaky outright rather than tracking
member access, which keeps the check a regex over source text with no new
dependency.
Review also suggested excluding *.test.tsx alongside *.test.ts. Skipped: the
repo has no .test.tsx files, and all four sibling checks in this file use
`.test.ts`. Diverging in one of them would read as an accident; if such files
appear, all four should change together.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
|
||
|
|
ff864ad3db |
feat(packs): sync system templates from packs/ instead of the frozen migration (#1390)
Phase 2b. The packs become the source of truth; the table stays the query surface, so the existing read path (one RLS query returning system + company + team templates) and every template id are untouched. pack_slug is the stable upsert key (migration 20260803230000, backfilled onto all 26 seeded rows). Matching on name instead would have meant that correcting a Swedish label looks like a new template, and because booking_template_usage.template_id is ON DELETE CASCADE, retiring the old row would silently wipe every company's "recently used" history for it. For the same reason a removed pack DEACTIVATES its row rather than deleting it: the read path already filters is_active, so it leaves the picker while usage history survives. The sync fails closed. A pack that will not parse or validate aborts the whole run with zero writes, because a database left in a state no commit of the repo describes is worse than a stale one. An empty catalogue is treated as a broken deploy (packs/ not bundled) rather than an instruction to retire every system template. Runs as a daily cron rather than at boot: boot-time work would have every serverless instance racing to write the same rows, and would re-apply a bad catalogue continuously instead of once a day where it is visible. Idempotent, so a database already matching the packs performs zero writes. Three database guards, each covered by tests/pg/booking-template-pack-slug.pg.test.ts: a partial unique index (one pack, one template), a format CHECK mirroring PACK_SLUG_RE so the database refuses what the loader would, and a CHECK keeping pack_slug off company templates, where it would shadow the pack it collides with. Verified the upgrade path locally the way the pg-upgrade job will run it: base schema, seeded fixture company with posted verifikat, an existing company template, then this migration alone. 26 slugs backfilled, company template intact, upgrade assertions pass. This is that job's first real migration. Payloads are spelled out rather than spread so the phantom-column guard can check them, and docker/crontab.* are regenerated for the new vercel.json entry. Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |