Replication · Flashcard

After a failover with async replication, the old leader comes back with writes the new leader never got. What happens to them, and why is that risky?

hard Availability

Answer

They are usually discarded when the old leader rejoins as a follower. That is dangerous if anything outside the database already saw them: for example, reused auto-increment ids can collide with keys a cache or another system stored for the lost rows. Promote the most up-to-date follower, and reconcile the discarded writes by hand.

Review this in your daily deck All cards in Replication