[ALERT]09:07 · Support reports that users can't open their profiles after the 09:00 deployment. API error rates look normal, but profile requests are failing.
09:00:02 INFO deployment PR#248 completed09:00:05 INFO backend healthcheck status=20009:03:11 WARN GET /api/profile status=40409:03:12 INFO GET /api/users/me status=20009:03:17 WARN GET /api/profile status=40409:03:20 WARN GET /api/profile status=404--- metrics: last 10 minutes ---GET /api/profile: 937 requests, 94% HTTP 404GET /api/users/me: 16 requests, 100% HTTP 200HTTP 500 errors: 0Support tickets: 23 profile access failures
Service Overview
- Service:
users-service - Stack: Node.js, Express, TypeScript (backend), React (frontend)
- Deployment: PR #248 refactored the profile routes.
- Recent Change: Renamed
/api/profileto/api/users/me.
System Invariants & Behavior
- The Express router is mounted at
/inapp.ts. - Authentication middleware and profile data handler (
getProfile) are working correctly. - The new endpoint
/api/users/mepasses health and smoke tests. - The frontend client (
src/client/profileApi.ts) still makes requests to/api/profile. - Client Cache Behavior: Existing browser sessions may continue executing cached JavaScript bundles after a deployment.
- Requirement: Users must be able to load their profile throughout the migration.
- The old endpoint
/api/profileno longer exists on the backend and returns HTTP 404.
Investigation Progress
Step 1 of 5(Mitigate)
Immediate Containment
It's 09:07 and users can't open their profiles after the 09:00 deployment. What do you do in the next two minutes to restore access before writing any code?
Be specific: what action, on what, and why it is safe.
Min 20 characters0 / 20 chars