Question

Difficulty: Very hardEnhancing Reliability and Disaster Recovery

An enterprise's primary workload runs in VPC-A within Account 1 (Production) in the us-east-1 region. The database layer uses an Amazon Aurora MySQL Global Database with the primary cluster in us-east-1 and a secondary cluster in us-west-2. Centralized DNS is managed in Account 2 (Shared Services) using a Route 53 private hosted zone (PHZ) named corp.internal. A CNAME record db.corp.internal points to the primary Aurora cluster endpoint in us-east-1. VPC-A is currently associated with this PHZ.

To improve disaster recovery, the enterprise is planning a pilot light strategy. During a region-wide outage in us-east-1, the application tier must failover to VPC-B within Account 1 in the us-west-2 region. The disaster recovery requirements specify a database Recovery Time Objective (RTO) of less than 15 minutes and a Recovery Point Objective (RPO) of less than 1 minute. Additionally, all DNS queries must remain strictly private, routing privately over the AWS network without using custom DNS servers or public Route 53 health check endpoints.

Which two actions should a solutions architect take to meet these requirements? (Select TWO.)

  1. Authorize the association of the corp.internal private hosted zone in Account 2 with VPC-B in Account 1 using cross-account VPC association authorization, then associate the zone in Account 1.Answer
  2. Implement Route 53 Application Recovery Controller (ARC) routing controls, create Route 53 health checks associated with the routing controls, and configure failover routing records in Account 2 for db.corp.internal pointing to the regional Aurora endpoints.Answer
  3. C
    Create a duplicate private hosted zone named corp.internal in Account 1 associated with VPC-B, and create a static CNAME record db.corp.internal pointing to the us-west-2 Aurora cluster endpoint.
  4. D
    Configure Route 53 failover routing records in Account 2 for db.corp.internal, and create standard Route 53 HTTP health checks pointing directly to the private IP addresses of the Aurora database endpoints to automate DNS failover.
  5. E
    Replace the Aurora Global Database with daily snapshot replication from us-east-1 to us-west-2, and write a Lambda function to restore the database from the latest snapshot in us-west-2 during failover.

Answer

Associate VPC-B with the centralized private hosted zone using cross-account authorization, and configure Route 53 Application Recovery Controller routing controls with failover routing records pointing to the regional database endpoints.
The correct architecture utilizes Aurora Global Database to achieve the required sub-minute RPO and multi-region failover speed. To manage DNS routing dynamically and privately, the centralized private hosted zone in Account 2 must be associated with the recovery VPC (VPC-B in Account 1) via cross-account VPC association authorization. Since standard Route 53 health checks cannot directly query private IP addresses in a VPC, Route 53 Application Recovery Controller (ARC) routing controls are implemented. Route 53 health checks are created to monitor the status of the ARC routing controls, allowing the failover routing policy in the private hosted zone to securely redirect the database endpoint resolution during failover without exposing queries or endpoints to the public internet.

Step-by-Step Solution

1
Configure cross-account private hosted zone association for the recovery VPC.
VPC-B in Account 1 is authorized and associated with the corp.internal private hosted zone in Account 2.
By default, private hosted zones are only accessible within the account they are created in. To allow VPC-B in Account 1 to resolve names in Account 2's PHZ, a cross-account authorization must be created in Account 2 using the CLI/API, and then accepted in Account 1.
2
Ensure database replication meets recovery objectives using Aurora Global Database.
Active-passive replication is established with sub-second replication latency between the primary and secondary regions.
Aurora Global Database utilizes physical storage-level replication to achieve an RPO of less than 1 second, directly supporting the business RPO constraint.
3
Configure Route 53 Application Recovery Controller (ARC) routing controls.
Routing controls are created, and corresponding Route 53 health checks are linked to the routing control states.
Because standard Route 53 health checks cannot access resources inside a private VPC, ARC routing controls are used as a control plane mechanism. The associated health checks monitor the control plane state rather than the data plane directly, keeping all monitoring traffic private.
4
Create DNS failover records in the private hosted zone in Account 2.
Failover records for db.corp.internal point to us-east-1 (primary) and us-west-2 (secondary), evaluated using the ARC-linked health checks.
During an outage, the database is failed over, and the ARC routing control state is updated to redirect application queries to the promoted secondary database cluster in us-west-2.

Key Concept

Multi-account private hosted zone routing and multi-region database failover using Route 53 ARC.
Rate this question