Question #731
A company needs to design a disaster recovery (DR) solution for its customer-facing web application. The application runs on a cluster of m5.xlarge Amazon EC2 instances managed by an Auto Scaling group across multiple Availability Zones and uses an Amazon RDS for PostgreSQL DB instance. During a disaster, the application must fail over to a secondary environment with an RPO of 30 seconds and an RTO of 10 minutes. Which solution meets these requirements MOST cost-effectively?
Use infrastructure as code (IaC) to deploy resources in the DR Region. Create a cross-Region read replica for the PostgreSQL DB. Configure AWS Backup to perform cross-Region backups of EC2 instances and the DB every 30 seconds using a cron job. Restore EC2 instances from backups during failover. Use Amazon Route 53 weighted routing policy to redirect traffic to the DR Region.
Use infrastructure as code (IaC) to deploy resources in the DR Region. Create a cross-Region read replica for the PostgreSQL DB. Implement AWS Elastic Disaster Recovery for continuous replication of EC2 instances to the DR Region. Maintain minimal EC2 capacity in the DR Region. Use Amazon Route 53 failover routing policy to automate traffic redirection. Scale the Auto Scaling group to full capacity during failover.
Deploy DR infrastructure using IaC. Use AWS Backup to create cross-Region backups of EC2 instances and the DB every 30 seconds. Manually restore backups to new instances in the DR Region during a disaster. Configure Route 53 geolocation routing to direct users to the DR Region. Adjust the Auto Scaling group size post-failover.
Use IaC to provision DR infrastructure. Set up an Amazon Aurora global database for cross-Region replication. Use AWS Elastic Disaster Recovery for EC2 replication. Run EC2 instances at full capacity in the DR Region. Route traffic using Route 53 latency-based routing. Trigger failover manually during disasters.
Explanation
Option B is correct because:
- Cross-Region RDS Read Replica: Ensures an RPO of 30 seconds by maintaining a near-real-time replica of the PostgreSQL database, which can be promoted quickly during a disaster.
- AWS Elastic Disaster Recovery (DRS): Provides continuous replication of EC2 instances, minimizing data loss (RPO) and enabling rapid recovery (RTO).
- Minimal EC2 Capacity in DR: Reduces costs by keeping only essential resources running, while Auto Scaling ensures rapid scaling to full capacity during failover.
- Route 53 Failover Routing: Automates traffic redirection to the DR Region, ensuring RTO compliance.
Other options fail because:
- A: AWS Backup cannot perform backups every 30 seconds, and restoring backups would exceed the RTO.
- C: Manual restoration from backups violates the RTO, and AWS Backup's frequency limitations conflict with the RPO.
- D: Aurora Global Database is incompatible with PostgreSQL, and running full EC2 capacity in DR is not cost-effective.
Key Points: Use cross-Region read replicas for RDS, continuous replication (e.g., AWS DRS) for EC2, minimal DR infrastructure, and automated DNS failover for cost-effective DR solutions with low RPO/RTO.
Answer
The correct answer is: B