Runbook Checklist Generator

# Runbook: Database Connection Pool Recovery
Last updated: 2026-06-09

## When to Use
When connection pool exhaustion alert fires (> 90% pool used for > 2 minutes)

## Prerequisites
- [ ] Access to production environment
- [ ] Monitoring dashboard open

## Steps
- [ ] Check current pool usage in Grafana
- [ ] Identify slow queries: SELECT * FROM pg_stat_activity WHERE wait_event_type='Lock'
- [ ] Kill blocking queries if safe: SELECT pg_terminate_backend(pid)
- [ ] Scale connection pool if needed: update POOL_SIZE env var
- [ ] Verify alert resolved in monitoring
- [ ] Notify team in #incidents channel

## Rollback
- [ ] Restore previous POOL_SIZE value
- [ ] Deploy previous configuration
- [ ] Verify service health

## Contacts & Escalation
DBA team: #database-ops | Platform: #platform-eng

## Related
- Monitoring: (link)
- Docs: (link)
- Post-mortem template: (link)