DBRaven
Consistency

Eventual Consistency

Intermediate

How eventual consistency models propagate updates across nodes, how convergence windows create read anomalies, how conflict resolution works, and when strong consistency is required instead.

Step 1 of 5

Consistent State: All Nodes Agree

In a consistent state, all replicas hold the same data. A read to any node returns the same value. Under eventual consistency, this is the steady state after all writes have propagated. Systems like Cassandra and DynamoDB work toward this state: but make no guarantee about when convergence will occur after a write.

Node A
consistent

user:42 → plan=free

Node B
consistent

user:42 → plan=free

Node C
consistent

user:42 → plan=free

All three nodes consistent: reads from any node return plan=free

Key Takeaways

  • Convergence is the eventual goal: but it is not instantaneous
  • Under no load and no failures, eventual consistency and strong consistency look identical
1 / 5