An international smart grid utility corporation hosts its core telemetry visualization application on AWS. The application tier runs on Amazon EC2 instances inside an Auto Scaling group behind an Application Load Balancer (ALB). The relational database tier is deployed as an Amazon RDS for MySQL DB instance. The company needs to design a cross-region disaster recovery (DR) solution that achieves a Recovery Time Objective (RTO) of 45 minutes and a Recovery Point Objective (RPO) of 15 minutes, while minimizing compute costs and operational overhead during normal operations. Which architecture should a solutions architect recommend to meet these requirements?
- Create a cross-region read replica of the RDS DB instance in the secondary Region. Deploy the Application Load Balancer and the Auto Scaling group in the secondary Region with the desired capacity set to 0. During failover, promote the read replica, scale up the Auto Scaling group, and update Route 53 DNS records.Answer
- BCreate a cross-region read replica of the RDS DB instance in the secondary Region. Configure the Application Load Balancer in the primary Region to automatically route write traffic to the read replica if the primary DB instance becomes unavailable.
- CDeploy the EC2 instances and ALB in the secondary Region. Configure Amazon Route 53 with a latency routing policy to dynamically direct all client traffic to the secondary Region to achieve automatic failover if the primary Region becomes unhealthy.
- DBack up the primary RDS DB instance to Amazon S3 daily, and copy the backups to the secondary Region using S3 Cross-Region Replication with an S3 Lifecycle policy that immediately transitions them to Amazon S3 Glacier Flexible Retrieval. Restore the database from the Glacier archives during a disaster.
Answer
Create a cross-region read replica of the RDS DB instance in the secondary Region. Deploy the Application Load Balancer and the Auto Scaling group in the secondary Region with the desired capacity set to 0. During failover, promote the read replica, scale up the Auto Scaling group, and update Route 53 DNS records.
The correct strategy uses the Pilot Light pattern to achieve the recovery targets while keeping costs minimal. An Amazon RDS cross-region read replica satisfies the 15-minute RPO through asynchronous replication, which usually lags by only a few seconds. Deploying the application tier with an Auto Scaling group set to 0 capacity ensures that you only pay for storage and metadata rather than active compute instances during normal operations. During failover, the replica is promoted, the Auto Scaling group is scaled up to handle the load, and Amazon Route 53 points traffic to the new ALB, easily completing the process within the 45-minute RTO.
Step-by-Step Solution
Key Concept
Disaster recovery strategies differ by Recovery Time Objective (RTO), Recovery Point Objective (RPO), and cost. A Pilot Light strategy keeps critical data replicated while keeping compute resources off or scaled to zero, minimizing cost while enabling recovery within minutes.