A company hosts a database that runs on an Amazon RDS instance deployed across multiple Availability Zones. The company periodically runs a script against the database to report new entries that are added. The script negatively affects the performance of a critical application. The company needs to improve application performance with minimal cost and the least operational overhead. Which solution will meet these requirements?

Creating an Amazon RDS read replica is the most effective solution because it offloads read-heavy reporting queries from the primary database instance. This prevents the reporting script from competing for resources with the critical application, thereby improving the application's performance without impacting the primary database's write operations. Read replicas are designed for this exact purpose, offering high availability and scalability for read workloads with minimal operational overhead and cost-efficiency. The other options are less suitable: Identifying the instance with fewest active connections is not applicable to a single RDS instance deployed across multiple AZs; it's still the same logical database. This approach also doesn't solve the resource contention issue. Manually exporting entries is inefficient, prone to human error, and does not meet the requirement for minimal operational overhead. Amazon ElastiCache is a caching service that can improve read performance for frequently accessed data, but it doesn't solve the problem of a long-running reporting script that needs to scan new entries, which might not be cacheable or frequently accessed in the same way. It also adds more operational complexity than a read replica for this specific use case.