supabase: API Gateway delays/timeouts (19-40s) across all operations - Similar to March 22 incident
Bug report
- I confirm this is a bug with Supabase, not with my own application.
- I confirm I have searched the Docs, GitHub Discussions, and Discord.
Describe the bug
Severe API Gateway delays and timeouts affecting all Supabase operations (auth, data queries, WebSocket connections). This appears identical to the March 22nd incident (“API Requests to projects being blocked at the API Gateway layer”) but is occurring again.
Specific issues:
- Auth token refresh requests taking 19.5s
- Data queries taking 20-40s
- Consistent 522 Cloudflare timeout errors
- WebSocket connections failing
- All operations experiencing gateway-level delays before reaching application code
To Reproduce
The issue is consistently reproducible across all Supabase operations:
- Any authentication request (login, token refresh) takes 19+ seconds or fails with 522
- Any data query takes 20-40 seconds or fails with 522
- WebSocket connections fail to establish
- API calls return Cloudflare 522 errors indicating gateway timeouts
Example error response:
{
'message': 'JSON could not be generated',
'code': 522,
'hint': 'Refer to full message for details',
'details': 'Connection timed out'
}
Expected behavior
- Auth requests should complete in milliseconds
- Data queries should complete in normal timeframes (< 1s)
- WebSocket connections should establish successfully
- No API Gateway timeouts or blocks
Screenshots
[See attached]
System information
- OS: macOS
- Browser: Chrome
- Version of supabase-js: [Your version]
- Infrastructure: Google Cloud Platform (App Engine)
Additional context
- Issue began in last 24-36 hours
- Affects all endpoints consistently
- Pattern matches March 22nd incident exactly (API Gateway blocking)
- Delays are consistent across different operation types (19-40s)
- Error patterns show this is happening at the API Gateway level before reaching application code
- Support ticket has been opened but no response yet
The systematic nature of the delays and the similarity to the March 22nd incident strongly suggest this is a recurrence of the API Gateway configuration issue that was previously resolved.
About this issue
- Original URL
- State: closed
- Created 2 years ago
- Comments: 22 (6 by maintainers)
We’re seeing this again as well. Third time this year. The reports section of our Supabase dashboard shows the elevated errors and response times starting yesterday. Love Supabase but we can’t keep dealing with this, let alone recommending it like we used to.