Production database stuck after AWS us-east-1 outage - Need help
roject ID: cffllblnrjbjaltgtoqfRegion: us-east-1 (Free tier)
Problem:
My production database got stuck during yesterday's AWS outage and hasn't recovered even though the outage is resolved. All connection attempts fail with:
FATAL: Unable to check out process from the pool due to timeout
What I've tried:
- ✅ Restarted pooler multiple times (15→20→15)
- ✅ Added ?pgbouncer=true to connection string
- ✅ Redeployed application
- ✅ Waited 12+ hours
- ✅ Submitted support ticket (SU-282150) - no response yet
- ❌ Nothing clears the stuck connections
My dev database in us-east-2 works fine - same code, different region. Proves this is infrastructure, not code.
What I need:
Free tier doesn't have "Restart Database" button. My production site has been down for 12+ hours.
Is there any way to:
1. Manually kill stuck connections?
2. Force a database restart on Free tier?
3. Get faster support response?
Should I just upgrade to Pro temporarily to get the restart button?
Any help would be greatly appreciated! 🙏
3 Replies
Restart button does exist for Free tier under general settings.
What is the status of the project itself from the homepage?
Update: Project restart did NOT fix the issue!
Timeline:
- Yesterday: AWS us-east-1 outage caused connection pool to fill
- Today: Outage resolved, but my database connections remained stuck
- Just now: Clicked "Restart project" button - restart completed successfully
- Current status: Project shows GREEN/Healthy, but connections STILL failing
Error persists:
FATAL: Unable to check out process from the pool due to timeout
This is critical:
- Project restart didn't clear stuck connections
- Dashboard shows "Healthy" but database is unusable
- Production site down for 15+ hours
- Support ticket (SU-282150) - no response
Questions:
1. Why would project restart NOT clear connection pool?
2. Is there a database-specific restart separate from project restart?
3. Do I need to upgrade to Pro to get proper database restart access?
This seems like a serious bug - project restart should have cleared all connections.
Please help! 🙏
Project restarts the DB too. Pro should not be involved in the issue itself.
Support is likely your avenue and 24 hours is not much for Free support these days.
Which connection string are you using (.co direct or .com supavisor)?