AWS Certified Database – Specialty — Question 75

A company is using Amazon Aurora PostgreSQL for the backend of its application. The system users are complaining that the responses are slow. A database specialist has determined that the queries to Aurora take longer during peak times. With the Amazon RDS Performance Insights dashboard, the load in the chart for average active sessions is often above the line that denotes maximum CPU usage and the wait state shows that most wait events are IO:XactSync.
What should the company do to resolve these performance issues?

Answer options

• A. Add an Aurora Replica to scale the read traffic.
• B. Scale up the DB instance class.
• C. Modify applications to commit transactions in batches.
• D. Modify applications to avoid conflicts by taking locks.

Correct answer: B

Explanation

Scaling up the DB instance class (option B) increases the resources available to the database, which can directly enhance performance during peak times. Adding an Aurora Replica (option A) primarily helps with read traffic, but the issue here is related to CPU and IO performance, not read capacity. Modifying applications to commit transactions in batches (option C) or to avoid conflicts by taking locks (option D) may help but do not address the underlying issue of CPU and resource constraints.