Definition

A read replica is a read-only copy of a database instance[[9]](https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_ReadRepl.html "Working with DB instance read replicas - Amazon Rela…"). The primary database handles writes; read replicas asynchronously replicate data from the primary and serve read queries.

Core concept

Read replicas distribute read load. When an application needs to scale beyond one server’s capacity, it can direct SELECT queries to replicas, offloading work from the primary.

Use cases

High-traffic systems with many read queries (web apps, reporting) use read replicas. For instance, a social network might route timeline fetches to replicas while writes (posting) go to primary.

Trade-offs

Replication adds eventual consistency trade-offs: replicas lag slightly behind the primary, so a recent write might not immediately show up on a replica. Also, replicas must periodically synchronize with the primary, which adds network and I/O overhead. But read throughput and availability improve.

Example

AWS RDS lets you create a read replica. The docs state: “A read replica is a read-only copy of a DB instance. You can reduce load on the primary by routing queries to the read replica, elastically scaling beyond the capacity of a single DB instance for read-heavy workloads”[[9]](https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_ReadRepl.html "Working with DB instance read replicas - Amazon Rela…").

References

AWS RDS docs on read replicas[[9]](https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_ReadRepl.html "Working with DB instance read replicas - Amazon Rela…").