Examkin

Question

Difficulty: Very hardHigh Availability and Disaster Recovery Design

A financial technology company is designing a new multi-region transaction processing platform. The platform must serve users in both the United States and Europe with minimal latency. Outbound transactions must connect to external payment processors via public APIs. The recovery objectives for the architecture are:
- Recovery Point Objective (RPO): Less than 1 second for database transactions.
- Recovery Time Objective (RTO): Less than 2 minutes for a complete regional outage.

The application tier runs on Amazon ECS on AWS Fargate in private subnets and connects to an Amazon Aurora MySQL database cluster. The network architecture must also ensure that the loss of a single Availability Zone (AZ) does not disrupt outbound connection capability to external payment gateways, and that internal API microservices deployed in separate VPCs within both regions can resolve the private DNS names of the database endpoints.

Which two options should a Solutions Architect select to meet these requirements?

  1. Deploy an Amazon Aurora Global Database with the primary writer cluster in us-east-1 and a secondary reader cluster in eu-west-1. Use Route 53 failover routing policy with health checks linked to the Application Load Balancers for traffic management, and execute a managed failover of the Aurora Global Database during a regional outage.Answer
  2. Deploy a NAT Gateway in each Availability Zone within each region, configuring the route tables of the private subnets to use the local NAT Gateway. Associate the database's Route 53 Private Hosted Zone with the application VPCs in both regions.Answer
  3. C
    Deploy a single NAT Gateway in the primary Availability Zone of each region to optimize outbound data transfer costs. Create a single Route 53 Private Hosted Zone in the primary database VPC, and configure the secondary region VPC to resolve endpoints using Transit Gateway routes.
  4. D
    Deploy Amazon Aurora multi-region read replicas using cross-region snapshot copy and restore. During a disaster recovery event, promote the replica database in eu-west-1, modify the database endpoint, and redeploy the Fargate tasks with updated environment variables.
  5. E
    Enable Aurora Auto Scaling replicas in the secondary region to automatically scale the reader endpoints to handle primary write traffic in active-active mode, using Route 53 Latency-based routing to direct write requests directly to the nearest regional database endpoint.

Answer

Deploy an Amazon Aurora Global Database with failover routing policies, deploy a NAT Gateway in each Availability Zone within each region, and associate the Route 53 Private Hosted Zone with the application VPCs in both regions.
The correct architecture uses Amazon Aurora Global Database to achieve the required sub-second database RPO and supports managed failover under the 2-minute RTO. Route 53 active-passive failover routing ensures traffic is correctly rerouted to the secondary region's ALBs during a regional disaster. High availability for outbound payment processing is achieved by deploying a NAT Gateway in each Availability Zone, isolating zone failures. Finally, associating the Route 53 Private Hosted Zone with the consumer VPCs in both regions allows proper DNS resolution of database endpoints across the network boundary.

Step-by-Step Solution

1
Select the multi-region database replication and failover mechanism.
Amazon Aurora Global Database provides physical replication lag of less than 1 second (meeting the RPO) and managed failover in under a minute (meeting the RTO).
Standard replication or snapshot recovery techniques do not provide the necessary replication speed or failover automation to meet the RPO and RTO constraints.
2
Ensure outbound HA by designing NAT Gateway placement.
Place a NAT Gateway in each Availability Zone so that an outage in one zone does not disrupt outbound egress from other zones.
Using a single NAT Gateway per region introduces a single point of failure (SPOF) for the entire VPC's outbound connections.
3
Configure DNS resolution for private database endpoints across VPCs.
Associate the Route 53 Private Hosted Zone of the database with all application VPCs in both regions.
Private Hosted Zones are only resolvable from VPCs that have been explicitly associated with them.
4
Determine global traffic routing policy.
Use Route 53 active-passive failover routing linked to Application Load Balancer health checks.
This setup ensures incoming client requests are redirected to the secondary region if the primary region's application tier is impaired.

Key Concept

High Availability and Disaster Recovery Design for Multi-Region Multi-AZ Architectures
Rate this question

Topics

High Availability and Disaster Recovery Design

More questions from this topic