A single-instance Amazon RDS database is experiencing CPU and I/O saturation during peak read traffic. Which two actions should the SysOps administrator take to reduce read load and improve performance? (Choose two.)

Adding one or more RDS read replicas directly addresses CPU and I/O saturation from read traffic by distributing the read load across multiple database instances. This offloads reads from the primary database, improving its performance and freeing up resources for writes. Caching frequently accessed data in Amazon ElastiCache for Memcached reduces the number of read requests that reach the database entirely. By serving these requests from a fast, in-memory cache, it significantly decreases the read load on the RDS instance, mitigating CPU and I/O bottlenecks. Moving to DynamoDB is a significant architectural change, not a quick fix for an existing RDS instance. Rehosting on EC2 would require managing the database yourself and doesn't inherently solve read saturation. Converting to Multi-AZ improves availability but does not offload read traffic or directly address performance issues from high read loads.