Soru

Zorluk: OrtaMulti-AZ, Multi-Region Architectures and Disaster Recovery (DR)

An insurance company runs its claims processing application in the us-east-1 Region. The application uses Amazon EC2 instances in an Auto Scaling group behind an Application Load Balancer, and an Amazon RDS for PostgreSQL DB instance. The company wants to design a disaster recovery (DR) plan in the us-west-2 Region. The DR strategy must achieve a Recovery Time Objective (RTO) of 1 hour and a Recovery Point Objective (RPO) of 15 minutes, while minimizing ongoing costs. Which combination of actions will meet these requirements? (Select TWO.)

  1. Configure a cross-region read replica of the RDS for PostgreSQL DB instance in the us-west-2 Region, and promote it to a standalone DB instance during failover.Cevap
  2. Pre-stage an Application Load Balancer and an Auto Scaling group in the us-west-2 Region with the minimum capacity set to 0, and configure the Auto Scaling group to launch instances using replicated Amazon Machine Images (AMIs) during a failover.Cevap
  3. C
    Configure an Amazon RDS Multi-AZ deployment spanning both the us-east-1 and us-west-2 Regions to handle database replication and failover automatically.
  4. D
    Deploy a copy of the compute tier in the us-west-2 Region with the Auto Scaling group configured to run at full capacity to handle the application workload immediately.
  5. E
    Schedule daily database backups to an Amazon S3 bucket, configure S3 Cross-Region Replication to the us-west-2 Region with S3 Glacier Flexible Retrieval as the storage class, and restore the DB instance from the backup during failover.

Cevap

To meet the requirements, the company should configure a cross-region read replica of the RDS for PostgreSQL DB instance in the secondary region and promote it to a standalone DB instance during failover, and pre-stage an Application Load Balancer and Auto Scaling group with a minimum capacity of 0 in the secondary region to launch instances using replicated AMIs when needed.
The correct combination of actions implements a cost-effective Pilot Light disaster recovery strategy. Creating an Amazon RDS cross-region read replica in the recovery region ensures that data is continuously and asynchronously replicated, maintaining a recovery point within minutes, which satisfies the 15-minute RPO. Pre-staging the Application Load Balancer and the Auto Scaling group with a minimum capacity of 0 ensures that no compute costs are incurred in the secondary region during normal operations. During a failover event, promoting the RDS read replica and scaling the compute instances up to the required capacity from replicated AMIs can be completed in under an hour, meeting the 1-hour RTO.

Adım Adım Çözüm

1
Evaluate the database replication strategy to satisfy the Recovery Point Objective (RPO) of 15 minutes.
Using an Amazon RDS cross-region read replica provides asynchronous replication with lag times typically under a few minutes, meeting the 15-minute RPO.
Traditional backup and restore methods from daily snapshots cannot guarantee an RPO of less than 15 minutes, and RDS Multi-AZ cannot span multiple regions.
2
Evaluate the compute tier configuration in the secondary region to satisfy the Recovery Time Objective (RTO) of 1 hour while minimizing costs.
Pre-staging an Application Load Balancer and an Auto Scaling group with a minimum capacity of 0 ensures zero compute costs during normal operations (Pilot Light strategy). During failover, scaling the Auto Scaling group and launching instances from replicated AMIs takes only a few minutes, well within the 1-hour RTO.
Running a full-scale warm standby or active-active compute tier would incur substantial ongoing costs, which violates the constraint to minimize costs.

Anahtar Kavram

Disaster Recovery (DR) strategies involve finding the optimal balance between Recovery Time Objective (RTO), Recovery Point Objective (RPO), and cost. A Pilot Light strategy provides a cost-effective mechanism by keeping databases updated via asynchronous replication (RDS read replica) and keeping compute resources scaled down or pre-configured but not running (Auto Scaling group capacity 0).
Bu soruyu puanla