AWS Certified Database – Specialty — Question 295

A company has deployed an application that uses an Amazon RDS for MySQL DB cluster. The DB cluster uses three read replicas. The primary DB instance is an
8XL-sized instance, and the read replicas are each XL-sized instances.
Users report that database queries are returning stale data. The replication lag indicates that the replicas are 5 minutes behind the primary DB instance. Status queries on the replicas show that the SQL_THREAD is 10 binlogs behind the IO_THREAD and that the IO_THREAD is 1 binlog behind the primary.
Which changes will reduce the lag? (Choose two.)

Answer options

• A. Deploy two additional read replicas matching the existing replica DB instance size.
• B. Migrate the primary DB instance to an Amazon Aurora MySQL DB cluster and add three Aurora Replicas.
• C. Move the read replicas to the same Availability Zone as the primary DB instance.
• D. Increase the instance size of the primary DB instance within the same instance class.
• E. Increase the instance size of the read replicas to the same size and class as the primary DB instance.

Correct answer: B, E

Explanation

Because the SQL_THREAD is lagging significantly behind the IO_THREAD, the bottleneck is the replica's capacity to process and write the incoming replication data, which is caused by the replicas being undersized (XL) compared to the primary (8XL). Scaling the replicas to match the primary's size (Option E) provides them with the necessary compute and I/O resources to keep pace. Alternatively, migrating to Amazon Aurora MySQL (Option B) eliminates traditional binlog-based replication lag because Aurora Replicas share the same virtualized storage volume as the primary instance, resulting in near-zero lag.