A healthcare telemedicine platform processes patient consultation records in its primary AWS Region, us-east-1. The architecture consists of a containerized application running on Amazon EKS and a transactional database on Amazon RDS for PostgreSQL. To comply with national health regulations, the platform must implement a disaster recovery (DR) strategy in a secondary Region, us-west-2, with a Recovery Time Objective (RTO) of 10 minutes and a Recovery Point Objective (RPO) of 5 minutes. The strategy must also allow the company to perform quarterly DR drills in an isolated environment without affecting the production workload. Which of the following strategies meets these requirements while minimizing costs during normal operations?
- Migrate the database to Amazon Aurora PostgreSQL-Compatible Edition and configure an Aurora Global Database with the primary cluster in us-east-1 and a single-node secondary cluster in us-west-2. Deploy a scaled-down Amazon EKS cluster in us-west-2 with the minimum node capacity required to run core services. Configure Amazon Route 53 with failover routing. For quarterly drills, clone the secondary Aurora database cluster using the database cloning feature, and run testing workloads against the clone and the secondary EKS cluster.Answer
- BMaintain Amazon RDS for PostgreSQL in us-east-1 and deploy a cross-region Read Replica in us-west-2. Create a standby Amazon EKS cluster in us-west-2 with a pilot light configuration where the worker node groups are configured with a desired capacity of zero. For quarterly drills, promote the RDS Read Replica to a standalone instance, scale up the EKS node group capacity, run test transactions directly on the promoted instance, and then rebuild the replication link back to us-east-1.
- CKeep the application on Amazon RDS for PostgreSQL in us-east-1 and configure a cross-region Read Replica in us-west-2. Deploy an active Amazon EKS cluster in us-west-2. Configure Amazon Route 53 with a Latency routing policy linked to Route 53 health checks. In the event of a primary region outage, rely on the Latency routing policy to automatically redirect write requests from the application to the read replica in us-west-2, allowing the database engine to handle replication and failover automatically.
- DConfigure AWS Backup to take daily snapshots of the Amazon RDS for PostgreSQL database and copy them to an Amazon S3 Glacier Flexible Retrieval vault in us-west-2. Store the Amazon EKS cluster configuration as an AWS CloudFormation template. During quarterly drills or a disaster, use the CloudFormation template to provision the EKS cluster, initiate a standard retrieval of the database snapshot from Glacier, restore the database to a new RDS instance, and update Route 53.
Answer
Migrate the database to Amazon Aurora PostgreSQL-Compatible Edition, configure an Aurora Global Database with a single-node secondary cluster, deploy a scaled-down Amazon EKS cluster in the secondary Region, and use Aurora's database cloning feature for isolated testing.
Migrating the database to Amazon Aurora PostgreSQL and setting up an Aurora Global Database ensures sub-second replication lag, meeting the 5-minute RPO. Deploying a scaled-down EKS cluster (Warm Standby) keeps the control plane active and instances warmed up, allowing fast scaling within 10 minutes (meeting the RTO). Aurora's fast database cloning feature allows creating an isolated copy of the secondary database cluster in the DR region, which can be modified during drills without affecting production data replication.
Step-by-Step Solution
Key Concept
Disaster Recovery strategies (Warm Standby vs. Pilot Light) and Cross-Region replication configurations using Amazon Aurora Global Databases and database cloning.