Skip to main content

Production DR Checklist

Use this checklist before production rollout. It does not replace application-specific continuity design.

Control Plane​

  • Operator, server, and web are highly available.
  • CRDs can be restored or recreated.
  • Webhook certificates are valid.
  • Server APIs have authentication and audit coverage.

Clusters And Storage​

  • Source and target clusters are Ready.
  • Object storage is resilient and monitored.
  • Both clusters can access the repository.
  • Credentials have a rotation process.

Application Protection​

  • Namespace and label scope are clear.
  • DataSync and ResourceSync policies match RPO.
  • Target StorageClass, IngressClass, Secrets, and CRDs are ready.
  • Standby resources do not serve production traffic.

Runbook​

  • Instance failover has been drilled.
  • Group failover has been drilled.
  • DNS, gateway, or traffic switching steps are defined.
  • Undo, reprotect, and cancel decision rules are defined.

Monitoring​

  • Backup failure.
  • Restore failure.
  • Sync delay.
  • Operation failure.
  • Repository unavailable.
  • Control-plane pod failures.