AWS Certified Solutions Architect – Professional — Question 740

A company's solutions architect is managing a learning platform that supports more than 1 million students. The company's business reporting team is experiencing slow performance while extracting large datasets from the database. The learning application is based on PHP and runs on Amazon EC2 instances that are in an Amazon EC2 Auto Scaling group behind an Application Load Balancer (ALB). Application data is stored in an Amazon S3 bucket and in an Amazon RDS for MySOL database. The ALB is the origin of an Amazon CloudFront distribution.

The solutions architect observes that slow read operations for SELECT queries are affecting the RDS for MySOL DB instance's CPU utilization. The solutions architect must find a scalable solution to improve the slow website performance with near-zero downtime. The solution also must provide automatic failover with no data loss.

Which solution will meet these requirements?

Answer options

• A. Create an incremental database backup by using Percona XtraBackup. Compress the backup files. Synchronize the backup files to Amazon S3. Restore the backup files from Amazon S3 to Amazon Aurora MySOL. Direct the application endpoint to the new Aurora DB instance.
• B. Convert the DB instance to a Multi-AZ deployment. Set the query_cache_type parameter on the database to zero. Increase the CloudFront caching TTL to reduce application server CPU utilization.
• C. Create an Amazon Aurora read replica from the DB instance. Wait until the read replica is synchronized with the source DB instance. Promote the read replica to a standalone DB cluster. Direct the application endpoint to the new Aurora DB instance.
• D. Create a read replica cluster on the DB instance. Use a Multi-AZ deployment. Synchronize the read replica with the primary DB instance. Promote the read replica as the primary DB instance.

Correct answer: C

Explanation

Creating an Amazon Aurora read replica directly from the RDS for MySQL instance allows for an easy, near-zero downtime migration to Aurora, which natively supports automatic failover and high performance. Once the replica is synchronized and promoted to a standalone cluster, pointing the application to it resolves the CPU issues while ensuring high availability. Other options, such as using backup/restore methods, would result in significant downtime and data loss, while Multi-AZ RDS deployments do not scale read performance as effectively as Aurora.