Trigger
Use this runbook when the database is unreachable, slow, overloaded, or producing errors that affect application workflows.
| Step | Action | Owner | Evidence | Status |
|---|
| 1 | Confirm whether application, API, or jobs are affected | Incident Commander | Impact summary | TBD |
| 2 | Check database server status | Operations | Server status screenshot | TBD |
| 3 | Check application database error logs | Backend | Sanitized log evidence | TBD |
| 4 | Identify recent migrations or heavy queries | Backend / Release Owner | Migration or query evidence | TBD |
| 5 | Pause high-risk jobs if they are increasing load | Operations | Queue status evidence | TBD |
| 6 | Confirm backup availability before risky recovery actions | Operations | Backup evidence | TBD |
Safety Rules
- Do not run destructive SQL without explicit approval.
- Do not truncate, delete, or alter production data without an approved recovery plan.
- Confirm backup availability before schema or data repair.
- If payment or wallet integrity may be affected, classify as SEV-1 until reconciled.
Recovery Criteria
| Check | Expected Result |
|---|
| Application DB connectivity | Healthy |
| API workflows | Working |
| Queue jobs | Processing normally |
| Error rate | Returned to baseline |
| Data integrity | No unresolved financial or tenant impact |