AWS Certified SysOps Administrator – Associate — Question 434
A company has an application that customers use to search for records on a website. The application's data is stored in an Amazon Aurora DB cluster. The application's usage varies by season and by day of the week.
The website's popularity is increasing, and the website is experiencing slower performance because of increased load on the DB cluster during periods of peak activity. The application logs show that the performance issues occur when users are searching for information. The same search is rarely performed multiple times.
A SysOps administrator must improve the performance of the platform by using a solution that maximizes resource efficiency.
Which solution will meet these requirements?
Answer options
- A. Deploy an Amazon ElastiCache for Redis cluster in front of the DB cluster. Modify the application to check the cache before the application issues new queries to the database. Add the results of any queries to the cache.
- B. Deploy an Aurora Replica for the DB cluster. Modify the application to use the reader endpoint for search operations. Use Aurora Auto Scaling to scale the number of replicas based on load.
- C. Use Provisioned IOPS on the storage volumes that support the DB cluster to improve performance sufficiently to support the peak load on the application.
- D. Increase the instance size in the DB cluster to a size that is sufficient to support the peak load on the application. Use Aurora Auto Scaling to scale the instance size based on load.
Correct answer: B
Explanation
Since search queries are rarely repeated, implementing Amazon ElastiCache would result in a very low cache hit rate, making it an inefficient solution. Deploying Aurora Replicas and directing the read-heavy search traffic to the reader endpoint with Aurora Auto Scaling dynamically adjusts to traffic demands and maximizes resource efficiency. Changing instance sizes automatically via Auto Scaling is not supported for vertical scaling in this manner, and altering storage configuration does not address the compute bottleneck caused by read queries.