AWS Certified Database – Specialty — Question 349

A bike rental company operates an application to track its bikes. The application receives location and condition data from bike sensors. The application also receives rental transaction data from the associated mobile app.
The application uses Amazon DynamoDB as its database layer. The company has configured DynamoDB with provisioned capacity set to 20% above the expected peak load of the application. On an average day, DynamoDB used 22 billion read capacity units (RCUs) and 60 billion write capacity units (WCUs). The application is running well. Usage changes smoothly over the course of the day and is generally shaped like a bell curve. The timing and magnitude of peaks vary based on the weather and season, but the general shape is consistent.
Which solution will provide the MOST cost optimization of the DynamoDB database layer?

Answer options

• A. Change the DynamoDB tables to use on-demand capacity.
• B. Use AWS Auto Scaling and configure time-based scaling.
• C. Enable DynamoDB capacity-based auto scaling.
• D. Enable DynamoDB Accelerator (DAX).

Correct answer: C

Explanation

DynamoDB capacity-based auto scaling is the most cost-effective choice because it dynamically adjusts provisioned throughput to match the gradual, bell-curve changes in traffic, even when peak times vary due to weather. On-demand capacity would be highly uneconomical for a predictable, high-volume workload of billions of RCUs and WCUs. Time-based scaling is not suitable since the peaks are weather-dependent rather than strictly schedule-based, and DAX only optimizes read performance rather than reducing overall write and read provisioned capacity costs.