What should the solutions architect recommend?
Build Amazon RDS read replicas.
Build the database as a larger instance type.
Build a database cache using Amazon ElastiCache.
Build a database cache using Amazon Elasticsearch Service (Amazon ES).
Explanations:
Amazon RDS read replicas can help improve read performance by distributing read requests, but they may not achieve sub-millisecond response times for all common read requests due to replication lag and the nature of the underlying SQL database.
Upgrading to a larger instance type may improve performance but does not guarantee sub-millisecond response times for read requests. This approach primarily increases resource availability rather than optimizing read performance specifically.
Building a database cache using Amazon ElastiCache allows for fast, in-memory data retrieval, achieving sub-millisecond response times for common read requests. ElastiCache can store frequently accessed data, which can tolerate some degree of staleness, making it suitable for the company’s needs.
While Amazon Elasticsearch Service (Amazon ES) is optimized for search and analytics, it is not primarily designed for fast caching of transactional data. Using it as a cache may not consistently provide sub-millisecond responses for common read requests, especially compared to dedicated caching solutions like ElastiCache.