dima/supabase
1
0 Fork 0
mirror of https://github.com/supabase/supabase.git synced 2026-10-09 03:15:06 +03:00
Code Issues Packages Projects Releases Wiki Activity
Files
fix/docs-e2e-vercel-wait-commit-status
supabase/apps/docs/content/guides/telemetry/debugging.mdx
T
Jordi Enricgithub-actions[bot]Jeremias Menichelli
 ac7e66f215 docs(telemetry): add a Debugging guide and log-query best practices (#47762) 
Two related docs changes for debugging Supabase, consolidated into one
PR.

## 1. New "Debugging" guide (`guides/telemetry/debugging`)

A methodology and entry page for debugging any Supabase issue, added
next to Logging in the Telemetry nav. It covers:

- **The debugging loop** — read the exact error, isolate the failing
layer, gather evidence, fix, verify.
- **The Supabase request stack** — the gateway fans out to PostgREST,
GoTrue, Storage, and Realtime as parallel services (not a chain), with
Postgres underneath. Explains why an API-layer permission or
empty-result error is usually a Postgres RLS/privilege issue.
- **Reading logs** — the narrow-query discipline (one source, bounded
window, widen along an anchor), linking the Logs Explorer guide rather
than duplicating query syntax.
- **Symptom to guide routing table** — maps each symptom to its layer
and the specific troubleshooting guide, acting as a front door to the
troubleshooting collection. All 47 links verified live.

This puts the debugging methodology in docs (owned and updatable)
instead of only in the agent skill.

## 2. Log-query best practices (`guides/telemetry/logs`)

Adds the three practices the existing Best practices list was missing,
all engine-agnostic: query one source at a time, follow a request across
sources with an anchor, and reference only confirmed field names.

## Follow-up (not in this PR)

The Logs Explorer now defaults to **ClickHouse** (single `logs` table,
`log_attributes` map), but `guides/telemetry/logs.mdx` and the two logs
troubleshooting guides still document the legacy **BigQuery** dialect
(`cross join unnest(metadata)`). They need a coordinated BigQuery to
ClickHouse migration pass:

- `guides/telemetry/logs.mdx`
-
`troubleshooting/how-to-interpret-and-explore-the-postgres-logs-OuCIOj.mdx`
-
`troubleshooting/discovering-and-interpreting-api-errors-in-the-logs-7xREI9.mdx`

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Added a new “Debugging” guide with a step-by-step workflow for
identifying where issues originate versus where they appear.
* Added a symptom-to-layer troubleshooting mapping and linked it from
Telemetry navigation.

* **Documentation**
* Updated Logs Explorer guidance to note its ClickHouse default and that
examples use legacy BigQuery syntax.
* Expanded Logs Explorer best practices, including querying one source
at a time, correlating with anchors, and using only confirmed field
names.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Jeremias Menichelli <jmenichelli@gmail.com>
2026-07-10 18:24:43 +02:00

79 lines
18 KiB
Plaintext
Raw Permalink Blame History

---
id: 'debugging'
title: 'Debugging'
description: 'Isolate and fix Supabase issues by reading the error, isolating the failing layer, and gathering evidence from logs.'
---
Debug by evidence, not by guessing. A Supabase error almost always surfaces at one layer but originates at another, so the fastest path to a fix is finding _where_ the problem is, not pattern-matching the symptom. Retrying a failed request rarely helps; isolating the layer does.
## The debugging loop
Work through these steps in order, skipping straight to a fix before you have evidence for the cause is the most common way to waste time on a bug.
1. **Reproduce the issue and read the error precisely.** Capture the exact status code, the error code, and the full message, not a paraphrase. A `401` is not a `403`; `PGRST002` is not `PGRST106`; a Postgres `SQLSTATE` such as `42501`, `42P01`, or `23505` points at the exact failure. The precise error is your strongest clue. If you're using `supabase-js`, remember that errors are **returned, not thrown**, check the `error` field in the `{ data, error }` response object. Make sure your code inspects `error` — a swallowed error is why many bugs look like "nothing happened".
2. **Locate the failing layer.** Use the request stack below. The status code and error code usually name the layer for you.
3. **Gather evidence for that layer.** Query its logs, run the security and performance advisors, and inspect the schema. Logs are the primary tool, and the layer you identified in the previous step tells you which log source to query. See [Reading logs](#reading-logs) below.
4. **Isolate the cause** using the troubleshooting guide for that layer (see [Find the guide for your symptom](#find-the-guide-for-your-symptom) below). Confirm your hypothesis against the evidence before you act. Most Supabase issues trace back to a small, known set of causes, and the guide explains how to tell them apart.
5. **Apply the fix, then verify.** Re-run the exact operation that failed and confirm it now succeeds, and that the corresponding log line is clean. A fix you haven't re-run is still a guess. If a couple of attempts don't resolve it, stop and gather more evidence rather than repeating the same change.
## The Supabase request stack
A request from a client passes through several layers before it reaches your data. Errors propagate upward, so the layer that _reports_ an error is often not the layer that _caused_ it.
Knowing the shape of the stack is what makes isolating the layer possible.
```
Client (supabase-js / SSR)
→ Edge / API gateway → edge_logs (HTTP status, routing, rate limits)
The gateway routes each request to ONE of these services. They run in parallel,
not as a chain:
├→ PostgREST (Data API) → postgrest_logs (low-signal; PGRST* evidence lives in edge_logs and postgres_logs)
├→ GoTrue (Auth) → auth_logs (login, JWT, OAuth, email)
├→ Storage API → storage_logs (uploads, object access)
└→ Realtime → realtime_logs (channels, presence, broadcast)
PostgREST, GoTrue, and Storage each reach the database independently:
→ Supavisor (connection pooler) → supavisor_logs (pooling, timeouts)
→ Postgres (SQL, RLS, triggers) → postgres_logs (SQLSTATE, RLS, functions)
```
Edge Functions sit outside this stack and log separately: `function_edge_logs` for the HTTP request to the function, and `function_logs` for `console` output from inside it.
<Admonition type="note">
A permission error or an unexpectedly empty result at the API layer is often a Postgres row-level security or privilege problem one layer down, though filters, authentication, query shape, or a stale schema cache can produce the same symptom. When in doubt, trace toward the database.
</Admonition>
## Reading logs
Once you know the layer, query that layer's log source directly rather than scanning everything. Pick one `source`, bound the time window, and select only the fields you need.
When a query comes up empty, widen along an anchor, such as a timestamp, request ID, or error code, to follow the same request into the adjacent source (for example from `edge_logs` into `postgres_logs`) instead of broadening into an unfiltered scan.
A wide, unfiltered query across every source buries the one line you need and, on paid projects, costs more in scanned data.
The [Logging guide](/docs/guides/telemetry/logs) covers the Logs Explorer, the available log sources, and how to write queries against them.
## Find the guide for your symptom
Match your symptom to a layer, confirm it against that layer's logs, then open the troubleshooting guide for the specific cause and fix. If a symptom could fit two layers (for example, an auth call failing with what looks like an RLS error), start with the layer closest to the database.
Supabase updates these troubleshooting guides continuously, so treat this table as a starting point rather than the final word: if your exact symptom isn't listed, search the [troubleshooting index](/docs/guides/troubleshooting) for the error string.
| Symptom / error | Layer → log source | Troubleshooting guides |
| ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Empty `data` array with rows present; wrong rows returned; UPDATE/DELETE affects 0 rows; `42501` permission denied; `service_role` still blocked; policy not matching | RLS & access → `postgres_logs` | [Empty select array](/docs/guides/troubleshooting/why-is-my-select-returning-an-empty-data-array-and-i-have-data-in-the-table-xvOPgx) · [service_role hits RLS](/docs/guides/troubleshooting/why-is-my-service-role-key-client-getting-rls-errors-or-not-returning-data-7_1K9z) · [Database API 42501](/docs/guides/troubleshooting/database-api-42501-errors) · [RLS Simplified](/docs/guides/troubleshooting/rls-simplified-BJTcS8) · [Deprecated RLS features](/docs/guides/troubleshooting/deprecated-rls-features-Pm77Zs) |
| `PGRST002`/`PGRST106`; "schema cache"; "could not find table/relationship"; new column or table not recognized; `42P01`; `520`; API returns nothing | Data API (PostgREST) → `edge_logs`, `postgres_logs` | [Refresh schema cache](/docs/guides/troubleshooting/refresh-postgrest-schema) · [PGRST002](/docs/guides/troubleshooting/postgrest-error-pgrst002-could-not-query-the-database-for-the-schema-cache-c396e9) · [New objects not recognized](/docs/guides/troubleshooting/postgrest-not-recognizing-new-columns-or-functions-bd75f5) · [42P01](/docs/guides/troubleshooting/resolving-42p01-relation-does-not-exist-error-W4_9-V) · [520 errors](/docs/guides/troubleshooting/fixing-520-errors-in-the-database-rest-api-Ur5-B2) · [API not returning](/docs/guides/troubleshooting/why-is-my-supabase-api-call-not-returning-PGzXw0) |
| Login/logout/session broken; JWT "invalid claim"/"missing sub"; cookies not sent; OAuth redirect wrong; OTP/magic-link expired; MFA/TOTP fails; auth `500`/`503`; emails not arriving | Auth → `auth_logs`, `postgres_logs` | [401 missing sub](/docs/guides/troubleshooting/auth-error-401-invalid-claim-missing-sub--AFwMR) · [500 auth errors](/docs/guides/troubleshooting/resolving-500-status-authentication-errors-7bU5U8) · [503 AuthRetryableFetchError](/docs/guides/troubleshooting/auth-error-503-authretryablefetcherror-51b88c) · [OTP expired](/docs/guides/troubleshooting/otp-verification-failures-token-has-expired-or-otp_expired-errors-5ee4d0) · [OAuth not redirecting](/docs/guides/troubleshooting/oauth-sign-in-isnt-redirecting-on-the-server-side-ShGMtr) · [No auth emails](/docs/guides/troubleshooting/not-receiving-auth-emails-from-the-supabase-project-OFSNzw) · [Next.js auth](/docs/guides/troubleshooting/how-do-you-troubleshoot-nextjs---supabase-auth-issues-riMCZV) |
| `statement timeout`; duplicate key or sequence error; trigger errors; slow `ALTER`; blocked queries; disk/memory/swap pressure; index size | Database (Postgres) → `postgres_logs` | [Statement timeout](/docs/guides/troubleshooting/canceling-statement-due-to-statement-timeout-581wFv) · [Duplicate key / sequence](/docs/guides/troubleshooting/inserting-into-sequenceserial-table-causes-duplicate-key-violates-unique-constraint-error-pi6DnC) · [Blocked queries](/docs/guides/troubleshooting/how-to-check-if-my-queries-are-being-blocked-by-other-queries-NSKtR1) · [Disk not shrinking](/docs/guides/troubleshooting/disk-size-not-shrinking-after-deleting-data-135390) · [Autovacuum stalled](/docs/guides/troubleshooting/autovacuum-stalled-due-to-inactive-replication-slot-d55aa2) · [High CPU](/docs/guides/troubleshooting/high-cpu-usage) |
| "too many connections"; "remaining connection slots"; `CONNECT_TIMEOUT`; pooler vs. direct connection; read-only transaction; `prepared statement already exists`; `no pg_hba.conf entry`; IPv4/IPv6; SASL/SCRAM | Connections & pooler → `supavisor_logs`, `postgres_logs` | [Too many connections](/docs/guides/troubleshooting/too-many-connections-for-database-postgres) · [Remaining slots](/docs/guides/troubleshooting/database-error-remaining-connection-slots-are-reserved-for-non-replication-superuser-connections-3V3nIb) · [Prepared statement exists](/docs/guides/troubleshooting/error-prepared-statement-xxx-already-exists-3laqeM) · [Read-only transaction](/docs/guides/troubleshooting/resolving-cannot-execute-update-in-a-read-only-transaction-on-transaction-pooler-connections-ef582c) · [CONNECT_TIMEOUT](/docs/guides/troubleshooting/troubleshooting-connect_timeout-or-hanging-queries-in-vercel-serverless-functions-775f92) · [Supavisor terminology](/docs/guides/troubleshooting/supavisor-and-connection-terminology-explained-9pr_ZO) |
| Edge Function `401`/`404`/`500`/`503`/`504`/`546`; CPU/memory/wall-clock limit hit; won't deploy; boot error; WebSocket drop; `esm.sh` import fails | Edge Functions → `function_edge_logs`, `function_logs` | [401](/docs/guides/troubleshooting/edge-function-401-error-response) · [500](/docs/guides/troubleshooting/edge-function-500-error-response) · [503 boot](/docs/guides/troubleshooting/edge-function-503-response) · [504](/docs/guides/troubleshooting/edge-function-504-error-response) · [546 resource limit](/docs/guides/troubleshooting/edge-function-546-error-response) · [Shutdown reasons](/docs/guides/troubleshooting/edge-function-shutdown-reasons-explained) · [Deploy fails](/docs/guides/troubleshooting/edge-function-fails-deploy) · [esm.sh import](/docs/guides/troubleshooting/importing-stripe-or-other-modules-from-esmsh-on-deno-edge-functions-throws-an-error-TmbB5p) |
| Realtime `TIMED_OUT`; `TooManyChannels`; silent disconnect; missed database changes; broadcast-from-DB warning; heartbeats | Realtime → `realtime_logs` | [TIMED_OUT](/docs/guides/troubleshooting/realtime-connections-timed_out-status) · [TooManyChannels](/docs/guides/troubleshooting/realtime-too-many-channels-error) · [Silent disconnects](/docs/guides/troubleshooting/realtime-handling-silent-disconnections-in-backgrounded-applications-592794) · [Broadcast warning](/docs/guides/troubleshooting/realtime-warn-sending-broadcast-message) · [Heartbeats](/docs/guides/troubleshooting/realtime-heartbeat-messages) · [Logger](/docs/guides/troubleshooting/realtime-debugging-with-logger) |
| Upload or list fails; public bucket inaccessible; `relation "objects" does not exist`; file size limit; folder or RLS issue | Storage → `storage_logs`, `postgres_logs` | [Public bucket upload/list](/docs/guides/troubleshooting/why-cant-i-uploadlistetc-my-public-bucket-Z6CmGt) · [403 RLS on upload](/docs/guides/troubleshooting/storage-error-403-forbidden-new-row-violates-row-level-security-policy-on-upload-a94384) · [relation objects does not exist](/docs/guides/troubleshooting/relation-objects-does-not-exist-error-during-storage-uploads-8f21f0) · [File size limits](/docs/guides/troubleshooting/upload-file-size-restrictions-Y4wQLT) · [Folder ops / hierarchical RLS](/docs/guides/troubleshooting/supabase-storage-inefficient-folder-operations-and-hierarchical-rls-challenges-b05a4d) |
| Webhook not firing; `pg_cron` job not running; `pg_net` queue stuck; `42501 ... http_request_queue` | Database jobs → `postgres_logs` | [Webhook debugging](/docs/guides/troubleshooting/webhook-debugging-guide-M8sk47) · [pg_cron debugging](/docs/guides/troubleshooting/pgcron-debugging-guide-n1KTaz) · [42501 http_request_queue](/docs/guides/troubleshooting/42501--permission-denied-for-table-httprequestqueue-KnozmQ) |
| Reading or querying logs; interpreting Postgres logs; finding API errors in logs; reading metrics | Diagnostics → any log source | [Logging guide](/docs/guides/telemetry/logs) · [Interpret Postgres logs](/docs/guides/troubleshooting/how-to-interpret-and-explore-the-postgres-logs-OuCIOj) · [API errors in logs](/docs/guides/troubleshooting/discovering-and-interpreting-api-errors-in-the-logs-7xREI9) · [Logging levels](/docs/guides/troubleshooting/understanding-postgresql-logging-levels-and-how-they-impact-your-project-KXiJRm) · [View database metrics](/docs/guides/troubleshooting/how-to-view-database-metrics-uqf2z_) |
For query performance and schema-design questions such as indexing, `EXPLAIN`, N+1 queries, or partitioning, see the [Postgres guides](/docs/guides/database).
Debugging is complete only once you've re-run the failing operation, confirmed it succeeds, and checked that the layer's logs show a clean result.
Reference in new issue View Git Blame Copy Permalink