You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Device and OS: Playwright tests using Chromium hitting API deployed on LF-04
App/package versions: LeapfrogAI API V9.0 and 10.0
Steps to reproduce
Run UI playwright tests (increase workers to 4 in playwright.config.ts) to see this bug quicker. If using 1 worker, the API is more stable and it can take many tries to re-create.
Expected result
Tests pass
Actual Result
Tests begin to fail suddenly once backend breaks and gives this error:
│ httpx.ConnectError: All connection attempts failed │
│ │
│ During handling of the above exception, another exception occurred: │
│ │
│ Traceback (most recent call last): │
│ File "/home/nonroot/.local/lib/python3.11/site-packages/leapfrogai_api/routers/supabase_session.py", line 60, in init_supabase_client │
│ await client.auth.set_session( │
│ File "/home/nonroot/.local/lib/python3.11/site-packages/gotrue/_async/gotrue_client.py", line 601, in set_session │
│ response = await self.get_user(access_token) │
│ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ │
│ File "/home/nonroot/.local/lib/python3.11/site-packages/gotrue/_async/gotrue_client.py", line 547, in get_user │
│ return await self._request("GET", "user", jwt=jwt, xform=parse_user_response) │
│ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ │
│ File "/home/nonroot/.local/lib/python3.11/site-packages/gotrue/_async/gotrue_base_api.py", line 118, in _request │
│ raise handle_exception(e) │
│ gotrue.errors.AuthRetryableError: All connection attempts failed
The tests will not pass until you restart the api-deployment pod as well as both supabase-kong pods.
It also appears to be triggered by Playwright E2E tests that interruupt active message streams (by clicking the pause button or reloading the page). This is not always the case, but the issue did seem to happen more when those tests were the first to start failing.
Visual Proof (screenshots, videos, text, etc)
Severity/Priority
Major - causes catastrophic failure of backend requiring reboot of multiple services
Additional Context
The text was updated successfully, but these errors were encountered:
Environment
Device and OS: Playwright tests using Chromium hitting API deployed on LF-04
App/package versions: LeapfrogAI API V9.0 and 10.0
Steps to reproduce
Expected result
Tests pass
Actual Result
Tests begin to fail suddenly once backend breaks and gives this error:
The tests will not pass until you restart the api-deployment pod as well as both supabase-kong pods.
It also appears to be triggered by Playwright E2E tests that interruupt active message streams (by clicking the pause button or reloading the page). This is not always the case, but the issue did seem to happen more when those tests were the first to start failing.
Visual Proof (screenshots, videos, text, etc)
Severity/Priority
Major - causes catastrophic failure of backend requiring reboot of multiple services
Additional Context
The text was updated successfully, but these errors were encountered: