A company runs a car rental reservation portal in a single AWS Region. The application uses an Amazon RDS for MySQL DB instance. The company needs to establish a disaster recovery (DR) strategy in a secondary Region. The strategy must meet a Recovery Point Objective (RPO) of 24 hours and a Recovery Time Objective (RTO) of 4 hours. Which solution meets these requirements at the lowest cost?
- ADeploy a warm standby database replica in the secondary Region with a continuously running, scaled-down RDS DB instance to guarantee immediate database failover.
- Configure automated daily snapshots of the RDS DB instance and copy them to the secondary Region. In the event of a disaster, use AWS CloudFormation to deploy the application resources and restore the database from the copied snapshot.Answer
- CCreate an RDS Read Replica in the secondary Region, and configure Amazon Route 53 latency routing to automatically promote the replica and redirect database write traffic during a failover.
- DConfigure daily database backups, store them in Amazon S3 Glacier Flexible Retrieval in the secondary Region, and use standard retrieval to restore the database in the event of a disaster.
Answer
Configure automated daily snapshots of the RDS DB instance and copy them to the secondary Region. In the event of a disaster, use AWS CloudFormation to deploy the application resources and restore the database from the copied snapshot.
The Backup and Restore strategy using daily cross-region RDS snapshots and AWS CloudFormation templates is the most cost-effective option because it incurs zero active database or compute charges in the secondary Region. The daily snapshots satisfy the 24-hour Recovery Point Objective (RPO), and restoring the database and infrastructure can be completed within the 4-hour Recovery Time Objective (RTO).
Step-by-Step Solution
Key Concept
Disaster Recovery (DR) strategies on AWS vary in cost and recovery times. Backup and Restore is the most cost-effective DR strategy for workloads that can tolerate an RTO and RPO of several hours, as it does not require running active compute or database instances in the recovery region.
Estimated Time:1m 0s