An application deployed on Amazon ECS uses an Amazon RDS for MySQL DB instance. The app issues far more reads than writes. During peak traffic, application performance drops and the DB instance’s ReadLatency metric in CloudWatch spikes suddenly. What change should the developer make to improve performance?

Introducing Amazon ElastiCache (either Redis or Memcached) is the most effective solution because it caches frequently accessed query results, significantly reducing the read load on the Amazon RDS for MySQL instance. This directly addresses the observed spike in ReadLatency during peak traffic. Scaling the ECS cluster by adding more container host instances would help if the bottleneck were the application servers themselves, but the problem explicitly states the DB instance's ReadLatency is spiking, indicating a database bottleneck, not an application server bottleneck. Increasing read capacity units (RCUs) is relevant for Amazon DynamoDB, not Amazon RDS for MySQL. Adjusting the ECS task definition to allocate more memory to tasks would help if the application itself were memory-constrained, but again, the problem points to database read latency as the issue.