Amazon Redshift multi-Region disaster recovery

This article explores three disaster recovery strategies for Amazon Redshift to protect analytics workloads against regional disruptions, with guidance on selecting the right approach based on RPO and RTO requirements.

• Active-Passive uses cross-Region snapshot replication with 15-20 min RPO and 1-2 hour RTO at low cost
• Active-Active runs full pipelines in both regions simultaneously for near-zero RPO and minute-scale RTO at high cost
• Hybrid applies different strategies to data tiers, balancing protection and cost for mixed criticality workloads
• Use multi-Region AWS KMS keys, infrastructure as code, and comprehensive monitoring across all strategies
• Test DR procedures quarterly to validate RTO/RPO targets and ensure runbooks remain current
• Replicate S3 data, AWS Glue Catalog, and DynamoDB state alongside Redshift snapshots for complete resilience

Selecting the right DR strategy requires assessing business impact of downtime, data backfill capabilities, regulatory obligations, and cost tolerance to match workload criticality with appropriate recovery objectives.