Read Replica

What it is

A read replica is a copy of the database used for reads.

The primary handles writes.

The replica receives the data afterward and answers part of the read traffic.

When it matters

It becomes relevant when the system has heavy read traffic and you do not want every query hitting the primary.

That helps you:

• reduce read pressure on the primary
• isolate heavier queries
• scale reads without changing the write path first

Common mistake

The classic mistake is treating a read replica as if it reflected a write immediately.

It often does not.

If replication is delayed, the answer can be stale for a while.

Short example

A user saves a profile change.

The write goes to the primary.

Right after that, the UI reloads the data from a read replica.

If replication has not caught up yet, the user sees the old value and assumes the write failed.

Why it helps

Read replicas are great for scaling reads.

But they only help when the product is honest about which flows can tolerate delay and which ones need to read from the primary.

A read replica is not just “a cheaper database.” It is scalable reads with a different consistency model.