Examkin

Soru

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

A financial services company runs a mission-critical web application on AWS. The application uses Amazon EC2 instances in an Auto Scaling group behind an Application Load Balancer (ALB) and an Amazon Aurora MySQL database. The company requires a cross-region disaster recovery (DR) solution with a Recovery Time Objective (RTO) of 15 minutes and a Recovery Point Objective (RPO) of 2 minutes. The architecture must minimize ongoing idle compute costs. Which solution best meets these requirements?

  1. Configure Amazon Aurora Global Database with the primary database cluster in the active Region and a secondary cluster in the recovery Region. Set up an Auto Scaling group in the recovery Region with the desired capacity set to zero, maintaining pre-configured launch templates and target groups. Create a cross-Region Amazon S3 replication pipeline for application state. Use Amazon Route 53 Application Recovery Controller (ARC) routing control to fail over by promoting the secondary Aurora cluster and scaling the Auto Scaling group to the required capacity.Cevap
  2. B
    Set up Amazon Aurora cross-Region Read Replicas from the active Region to the recovery Region. Configure an Application Load Balancer in the recovery Region pointing to EC2 instances in a scaled-down Warm Standby state. Rely on Amazon Route 53 Latency routing with health checks to automatically promote the cross-Region Read Replica to primary and route traffic to the recovery Region if the active Region becomes unhealthy.
  3. C
    Establish a Pilot Light disaster recovery environment in the recovery Region. Schedule Aurora database snapshots every 15 minutes and replicate them to the recovery Region. Deploy an Auto Scaling group scaled to zero in the recovery Region. During a failover event, restore the database from the latest replicated snapshot, scale up the EC2 instances, and update Route 53 failover records.
  4. D
    Configure a Backup and Restore strategy by exporting Aurora database backups every 2 minutes to Amazon S3, and use S3 cross-region replication to copy these files to the recovery Region where they are transitioned to Amazon S3 Glacier Flexible Retrieval. Set up an Auto Scaling group in the recovery Region. During failover, use Standard Retrieval to restore the database in the recovery Region.

Cevap

Configure Amazon Aurora Global Database with the primary database cluster in the active Region and a secondary cluster in the recovery Region. Set up an Auto Scaling group in the recovery Region with the desired capacity set to zero, maintaining pre-configured launch templates and target groups. Create a cross-Region Amazon S3 replication pipeline for application state. Use Amazon Route 53 Application Recovery Controller (ARC) routing control to fail over by promoting the secondary Aurora cluster and scaling the Auto Scaling group to the required capacity.
The configuration using Amazon Aurora Global Database satisfies the 2-minute RPO due to its typical sub-second replication latency. Keeping the recovery Region's EC2 Auto Scaling group capacity at zero minimizes compute costs when idle, conforming to the Pilot Light architecture pattern. Since launch templates and target groups are pre-configured, the EC2 instances can scale up quickly, and promoting the Aurora secondary cluster takes less than a minute, allowing the entire recovery workflow to complete well within the 15-minute RTO. Route 53 ARC routing controls provide the mechanism to safely orchestrate this cross-region traffic shift.

Adım Adım Çözüm

1
Evaluate the RPO requirement of 2 minutes.
Identify that backup-and-restore or database snapshot strategies with intervals greater than 2 minutes are insufficient. Aurora Global Database replication is asynchronous and typically completes in under a second, meeting the RPO.
Choosing the correct replication technology is necessary to ensure data loss is kept within the RPO window.
2
Evaluate the RTO requirement of 15 minutes and the cost constraint.
Determine that compute resources in the recovery Region should be kept shut down or scaled to zero when idle (Pilot Light) to minimize cost, but must be launchable within minutes to meet the RTO.
Maintaining fully running EC2 instances (Warm Standby or Active-Active) would violate the cost optimization goal, while recreating infrastructure from scratch would violate the RTO.
3
Analyze database promotion and routing mechanism during failover.
Identify that promoting an Aurora Global Database secondary cluster is a fast, coordinated action (typically under a minute). Route 53 Application Recovery Controller (ARC) provides the necessary health checks and routing controls to coordinate the shift of application traffic safely.
Automated cross-region database promotion does not happen natively via Route 53 DNS policies; it requires orchestration or manual trigger, which Route 53 ARC routing controls support.

Anahtar Kavram

Selecting the appropriate disaster recovery strategy (Pilot Light vs Warm Standby vs Backup & Restore) and technology stack (Aurora Global Database, Route 53 ARC) to meet strict RTO/RPO requirements while optimizing costs.
Tahmini Süre:3m 0s
Bu soruyu puanla

Konular

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

Bu konudan daha fazla soru