NoSQL Consistency Models

Eventual consistency: updates propagate asynchronously, so replicas may briefly disagree, but given time without new writes they all converge. It favors availability and partition tolerance over immediate consistency.

Strong consistency: after a write completes, all subsequent reads see that value. It simplifies reasoning but requires coordination across replicas, increasing latency and reducing availability during partitions.

Read-your-writes: a session guarantee ensuring that once you write a value, your later reads reflect it, even if other clients might briefly see stale data. It improves perceived consistency for the user who made the change.

Tunable consistency: by setting read and write quorums (e.g., QUORUM, ONE, ALL), you balance consistency and availability per operation. If reads + writes replicas exceed the total, strong consistency is achieved.

Monotonic reads: guarantees that successive reads by a client do not go backwards in time. Without it, hitting different replicas could show a newer value then an older one, which confuses users.