A developer is creating a three-tier web application that must handle at least 5,000 requests per minute. The web tier must remain fully stateless, but user session state must be preserved. Which approach externalizes session data while keeping latency as low as possible?

An Amazon ElastiCache (Memcached) cluster provides an in-memory key-value store, offering extremely low-latency read and write operations, which is ideal for session management in a high-throughput, stateless web application. Storing session data in Memcached externalizes it from the web tier, allowing the web servers to remain stateless and scale independently. Storing session data in an Amazon RDS instance would introduce higher latency compared to an in-memory cache and could become a bottleneck for 5,000 requests per minute. A shared file system would also have higher latency and be less performant than an in-memory cache. While Amazon DynamoDB offers low-latency access, ElastiCache (Memcached) is specifically designed for caching and often provides even lower latency for simple key-value lookups, making it a more optimal choice for session state.