Multi-region and disaster recovery

Comments

• The dashed line is a contingency path: it is used only when the primary region fails.
• Aurora Global Database replicates data across regions, typically with less than one second of latency. Promoting the secondary region can be manual or automated.
• ECS in the secondary region can run on standby with fewer tasks to reduce cost, but it must be tested regularly.
• A recovery plan that has never been tested is not a plan. Run failure drills regularly to validate your RTO and RPO.

Information flow

1. Amazon Route 53

   The user resolves the domain with Route 53, which has a failover routing policy.

2. Amazon ECS (us-east-1)

   While the primary region passes its health checks, Route 53 sends traffic to ECS in that region.

3. Route 53 (contingency)

   If the primary region fails, Route 53 redirects traffic to the secondary region.

4. Aurora Global Database

   Aurora replicates data from the primary to the secondary region so recovery starts with current data.