Question #1163
A social media analytics platform uses Amazon Aurora as its primary database. During peak analytics processing times, the application experiences latency issues. Amazon CloudWatch metrics show significant spikes in Read IOPS and CPU Utilization when analytics queries are executed. What is the MOST cost-effective solution?
Migrate the analytics processing workload to Amazon Redshift.
Route analytics queries to an Aurora Replica using the reader endpoint.
Increase the Provisioned IOPS capacity for the Aurora database.
Upgrade the Aurora database instance to a larger instance class.
Explanation
The correct answer is B. Aurora Replicas allow read-only queries to be offloaded from the primary instance, distributing the read workload. This reduces pressure on the primary instance's Read IOPS and CPU, addressing latency during peak times without incurring the costs of scaling up (Option D) or increasing Provisioned IOPS (Option C). Migrating to Redshift (Option A) introduces complexity and costs for a workload Aurora can handle. Aurora's reader endpoint automatically routes queries to available replicas, making this the most cost-effective solution.
Key Points:
- Aurora Replicas handle read-only queries, reducing primary instance load.
- Reader endpoint distributes traffic across replicas.
- Scaling via read replicas is more cost-effective than vertical scaling or IOPS increases.
- Analytics workloads are typically read-heavy, making read scaling ideal.
Answer
The correct answer is: B