Use this when service recovery takes hours after a failure or you have backups but never actually tested restoring them. Describe your architecture and SLA requirements to get RPO/RTO targets, recovery procedures by failure type, and failover automation documented as a runbook.