A health insurance provider is migrating its transactional claims processing application to AWS. The application database requires a MySQL-compatible relational engine. The solutions architect must design a highly resilient database architecture that spans two AWS Regions in an active-passive setup. The design must achieve a Recovery Time Objective (RTO) of less than 15 minutes and a Recovery Point Objective (RPO) of less than 1 minute for a regional disaster. Furthermore, the database must support automatic failover in the primary region within 60 seconds if an Availability Zone fails.
Which configuration meets these requirements with the least operational complexity?
- Deploy an Amazon Aurora MySQL Multi-AZ cluster in the primary region, and configure an Amazon Aurora Global Database with a secondary cluster in the secondary region.Answer
- BDeploy an Amazon RDS for MySQL Multi-AZ DB instance in the primary region, and configure a cross-region read replica in the secondary region to act as the primary failover target using custom failover scripts.
- CDeploy a single Amazon RDS for MySQL DB instance in the primary region, and configure AWS Backup to take hourly database snapshots, copying them to the secondary region to restore a new DB instance during a disaster.
- DDeploy two independent Amazon RDS for MySQL Multi-AZ DB instances, one in the primary region and one in the secondary region, and use an Amazon Route 53 latency routing policy to synchronize data write operations between them.
Answer
Deploying an Amazon Aurora MySQL Multi-AZ cluster in the primary region and using Amazon Aurora Global Database to create a secondary cluster in the secondary region meets all constraints with the lowest operational overhead.
Deploying an Amazon Aurora MySQL Multi-AZ cluster in the primary region and using Amazon Aurora Global Database to create a secondary cluster in the secondary region meets all constraints with the lowest operational overhead. Aurora Multi-AZ clusters use shared storage and automatically fail over to a reader instance in another Availability Zone within 30 seconds, satisfying the local high availability requirements. For disaster recovery, Aurora Global Databases use storage-based, physical replication that operates asynchronously but has latency typically under 1 second (satisfying the Recovery Point Objective of less than 1 minute). In a disaster event, the secondary cluster can be promoted to a standalone primary cluster with read-write capabilities in under 10 minutes (satisfying the Recovery Time Objective of less than 15 minutes). This setup is managed natively by AWS, minimizing operational complexity.
Step-by-Step Solution
Key Concept
Amazon Aurora Global Database provides low-latency cross-region storage replication (RPO < 1 second) and fast managed failover/promotion (RTO < 10 minutes) for multi-region resilience, while Aurora Multi-AZ clusters handle rapid, automated local failovers (RTO < 30 seconds).