Question

Difficulty: HardHigh Availability and Disaster Recovery Design

An energy utility company is designing a disaster recovery architecture for its smart grid control system. The application's primary deployment is in us-west-2 and consists of microservices running on Amazon EC2 instances inside private subnets, outbound traffic routed via NAT Gateways, and a backend Amazon Aurora MySQL database. The company needs to design a disaster recovery solution in us-east-1 that achieves a Recovery Time Objective (RTO) of 10 minutes and a Recovery Point Objective (RPO) of 1 minute, while minimizing cost and avoiding active-active write workloads. Which TWO configurations should a Solutions Architect implement to meet these requirements?

  1. Create an Amazon Aurora Global Database with the primary database cluster in us-west-2 and a secondary database cluster in us-east-1 to replicate data with sub-second replication latency.Answer
  2. Deploy the application servers in us-east-1 using Auto Scaling groups configured with minimum running instances, and configure Route 53 active-passive failover routing to redirect traffic to the secondary region when the primary region is unhealthy.Answer
  3. C
    Implement hourly database snapshots in us-west-2, copy them to us-east-1 via cross-region snapshot copy, and restore them to a new database cluster during a failover event.
  4. D
    Deploy a single NAT Gateway in a single Availability Zone in us-east-1 to route outbound telemetry verification requests from all recovery private subnets to the internet.
  5. E
    Configure the application servers in the secondary region to write directly to the read-only standby instance of a standard Amazon RDS Multi-AZ deployment located in the primary region.

Answer

Creating an Amazon Aurora Global Database to achieve sub-second data replication and deploying the application servers in us-east-1 in a warm standby state with Route 53 active-passive failover routing.
To meet the 1-minute RPO and 10-minute RTO constraints while avoiding active-active complexity, a Solutions Architect should implement an Amazon Aurora Global Database for sub-second database replication across regions, and deploy a warm standby application stack in us-east-1. The warm standby uses Auto Scaling groups with minimum capacity to control cost, and Route 53 active-passive failover routing automatically or manually redirects traffic during a failover event.

Step-by-Step Solution

1
Analyze RPO requirements to determine the database replication strategy.
An RPO of 1 minute requires continuous or near-continuous data replication. Amazon Aurora Global Database uses storage-level replication to achieve sub-second cross-region replication lag, satisfying the 1-minute RPO.
Traditional backup and restore methods or snapshot replication are hourly or daily processes that violate the 1-minute data loss threshold.
2
Analyze RTO and cost requirements to determine the compute failover strategy.
An RTO of 10 minutes requires application infrastructure to be pre-provisioned but can be scaled down (warm standby) to minimize costs. Route 53 active-passive failover routing routes traffic to the recovery region when health checks fail.
Starting instances from scratch or redeploying infrastructure during a disaster would likely exceed the 10-minute recovery window, whereas active-active setups are too costly and complex.
3
Verify high availability constraints in the networking and compute paths of the recovery region.
Ensure NAT Gateways are deployed in each Availability Zone where EC2 instances reside to avoid a single point of failure.
A single NAT Gateway shared across multiple Availability Zones creates a non-redundant network path that can cause total outbound traffic loss if that Availability Zone goes down.

Key Concept

Designing multi-region disaster recovery using warm standby compute configurations and low-latency storage replication mechanisms like Aurora Global Database.
Rate this question