Multi-Region Disaster Recovery Test Checklist

Retrospective: this article looks back at events from December 2021, written in 2026 with the benefit of hindsight.

Use this checklist to plan and run a multi-region disaster recovery test for an AWS workload.

Before the test

• Workload owner, RTO and RPO targets documented.
• DR strategy chosen (backup/restore, pilot light, warm standby, active-active).
• Data replication status verified (replication lag, last successful backup copy).
• Infrastructure as code for the secondary region is current.
• KMS keys, secrets and parameters exist in the secondary region.
• DNS failover mechanism documented and tested in a non-production domain.
• Test scope, success criteria and rollback plan agreed.
• Stakeholders and customers informed if user-facing impact is possible.

During the test

• Start time recorded.
• Failover triggered using the documented procedure, by someone other than its author.
• Application health checks pass in the secondary region.
• Data integrity verified (recent transactions present).
• Time to restore service recorded.
• Monitoring and alerting work in the secondary region.
• Support and communication steps exercised.

Failback

• Failback procedure executed and timed.
• Data changes made during failover reconciled.

After the test

• Actual RTO and RPO compared with targets.
• Gaps documented with owners and dates.
• Runbook updated.
• Results reported to leadership.

Frequency

• Critical workloads: at least annually; after major architecture changes.