Managing Database Recovery

Use this skill to design or review recovery plans for data loss, corruption, or operational incidents.

Clarify First

• Likely incident type such as accidental deletion, corruption, outage, or region loss.
• Recovery time objective and recovery point objective.
• Backup types available and how recently they were validated.
• Whether the system can tolerate partial read-only or degraded modes.
• Which teams and systems depend on the restore sequence.

Recovery Principles

• Backups are only useful if restores are tested.
• Restoration order matters for dependent services.
• Snapshot age and data-loss tolerance must be explicit.
• A rushed restore without validation can compound the incident.
• Communication and decision points belong in the plan, not only technical steps.

Good Output

• Incident triage checklist.
• Restore decision tree.
• Validation steps before reopening traffic.
• Risks, assumptions, and fallback options.
• Questions to resolve before an actual incident.

Common Gaps

• No tested restore path.
• No clarity on point-in-time recovery capabilities.
• No post-restore verification plan.
• Confusing replication with recoverability.