Examkin

Question

Difficulty: HardHigh Availability and Disaster Recovery Design

A financial technology company is deploying a credit card payment gateway on AWS. The application must be highly available and resilient to regional disasters, with a target Recovery Time Objective (RTO) of less than 10 minutes and a Recovery Point Objective (RPO) of less than 1 minute. The primary infrastructure is located in us-east-1, and a secondary warm standby infrastructure is deployed in us-west-2. The application tier consists of Amazon EC2 instances in Auto Scaling groups behind Application Load Balancers. The database layer uses Amazon Aurora PostgreSQL. Outbound API calls to external payment networks must flow through NAT Gateways. Internal microservice endpoints are resolved using a Route 53 private hosted zone.

Which combination of actions should the Solutions Architect take to meet these requirements? (Select TWO.)

  1. Configure Amazon Aurora Global Database with the primary cluster in us-east-1 and a secondary cluster in us-west-2 containing an active Aurora Serverless v2 reader instance.Answer
  2. Create a Route 53 failover routing policy for the public DNS record, directing traffic to the us-east-1 load balancer as primary and the us-west-2 load balancer as secondary, and associate it with an active Route 53 health check.Answer
  3. C
    Configure a daily snapshot lifecycle in AWS Backup for the primary Aurora cluster in us-east-1, replicate the snapshots to us-west-2 every 12 hours, and restore the cluster in us-west-2 during a disaster event.
  4. D
    Associate the Route 53 private hosted zone only with the VPC in us-east-1, and rely on cross-region VPC peering to route DNS queries from the us-west-2 VPC.
  5. E
    Deploy a single NAT Gateway in us-east-1 and update the route tables of the private subnets in both us-east-1 and us-west-2 to route all outbound internet traffic through this NAT Gateway to minimize ongoing NAT hourly charges.

Answer

Configure Amazon Aurora Global Database with the primary cluster in the primary region and a secondary cluster in the standby region containing an active Aurora Serverless v2 reader instance, and create a Route 53 failover routing policy for the public DNS record directing traffic to the primary load balancer as primary and the secondary load balancer as secondary with an active health check.
The option proposing Amazon Aurora Global Database is correct because it uses dedicated physical replication to copy data from the primary to the secondary region with sub-second replication lag, meeting the 1-minute RPO. The secondary cluster can be promoted in less than a minute, satisfying the 10-minute RTO. The option proposing Route 53 failover routing is correct because it automatically detects failures in the primary region's ALB and dynamically redirects client traffic to the secondary region's ALB, keeping recovery times well within the 10-minute RTO.

Step-by-Step Solution

1
Determine the required cross-region database replication strategy to satisfy the 1-minute RPO requirement.
Identify Amazon Aurora Global Database as the correct solution because it replicates data asynchronously with sub-second lag, ensuring compliance with the RPO.
Alternative backup-based replication strategies fail the 1-minute RPO limit.
2
Design the failover mechanism to meet the 10-minute RTO requirement.
Implement a Route 53 failover routing policy linked to a health check that monitors the primary endpoint and automatically switches traffic to the secondary region.
Automated DNS failover minimizes manual intervention and keeps recovery time within the 10-minute constraint.
3
Identify high-availability gaps in the network architecture.
Ensure NAT Gateways are deployed redundantly in each Availability Zone and region, and associate the Route 53 private hosted zone with the VPCs in both regions.
Single NAT Gateways or unassociated private hosted zones introduce single points of failure and routing outages during disaster recovery execution.

Key Concept

Multi-Region Active-Passive Disaster Recovery
Rate this question

Topics

High Availability and Disaster Recovery Design

More questions from this topic