A reporting process that previously completed in 15 minutes now takes an hour. The application runs on EC2 instances and reads data from an Amazon RDS for MySQL instance. CloudWatch shows consistently high Read IOPS on the RDS instance, even when reports are not running. The administrator must improve both performance and availability of the database. Which solution meets these goals?

Creating an RDS read replica offloads read-heavy operations, like reporting, from the primary database. This reduces the Read IOPS on the primary instance, improving its performance for transactional workloads. By directing the reporting job to the read replica's reader endpoint, the primary instance remains dedicated to writes and critical application reads. This also enhances availability as the primary instance is less burdened and the read replica provides a separate endpoint for reporting, preventing resource contention. Deploying ElastiCache would require significant application changes to manage caching logic and might not be suitable for complex reporting queries. CloudFront is a CDN for static and dynamic web content, not a database proxy. Increasing the RDS instance class (scaling up) would improve performance but wouldn't address the high Read IOPS on the primary instance during reporting, nor would it inherently improve availability by separating workloads.