AWS Certified Database – Specialty — Question 287

A financial services company runs an on-premises MySQL database for a critical application. The company is dissatisfied with its current database disaster recovery (DR) solution. The application experiences a significant amount of downtime whenever the database fails over to its DR facility. The application also experiences slower response times when reports are processed on the same database. To minimize the downtime in DR situations, the company has decided to migrate the database to AWS. The company requires a solution that is highly available and the most cost-effective.
Which solution meets these requirements?

Answer options

• A. Create an Amazon RDS for MySQL Multi-AZ DB instance and configure a read replica in a different Availability Zone. Configure the application to reference the replica instance endpoint and report queries to reference the primary DB instance endpoint.
• B. Create an Amazon RDS for MySQL Multi-AZ DB instance and configure a read replica in a different Availability Zone. Configure the application to reference the primary DB instance endpoint and report queries to reference the replica instance endpoint.
• C. Create an Amazon Aurora DB cluster and configure an Aurora Replica in a different Availability Zone. Configure the application to reference the cluster endpoint and report queries to reference the reader endpoint.
• D. Create an Amazon Aurora DB cluster and configure an Aurora Replica in a different Availability Zone. Configure the application to reference the primary DB instance endpoint and report queries to reference the replica instance endpoint.

Correct answer: C

Explanation

Amazon Aurora offers faster failover times (usually under 30 seconds) and higher availability than Amazon RDS for MySQL, addressing the need to minimize DR downtime. By utilizing an Amazon Aurora DB cluster, the primary application can write to the cluster (writer) endpoint while reporting queries are offloaded to the reader endpoint, which automatically load-balances across Aurora Replicas. This setup is highly cost-effective as the replica serves as both a failover target and a read-scaling instance, unlike RDS Multi-AZ which requires a dedicated passive standby.