What is Eventual Consistency?
A consistency model where replicas may briefly disagree after a write but converge to the same value once updates propagate.
Definition
Under eventual consistency, a write is acknowledged before every replica has applied it, so a read that lands on a lagging node can return stale data — but absent new writes, all replicas converge. Amazon DynamoDB, Cassandra, and DNS all embrace this model because it keeps systems available and fast during network partitions, the trade-off the CAP theorem makes explicit. Engineering around it means designing for staleness: read-your-own-writes sessions, version vectors or last-write-wins for conflict resolution, and idempotent consumers. Strong consistency remains available in these systems, but at higher latency and reduced availability.