Hinted Handoff
TechniqueDistributed SystemsSeniorStaffUpdated
If a target replica is down, a healthy node stores the write locally with a 'hint' and forwards it once the target recovers. Keeps write availability high.
When to Use Hinted Handoff
Leaderless or multi-master systems where short replica outages should not fail writes. Pairs with anti-entropy for durability if the hint-holder also dies.
In Practice
Cassandra hinted handoff, DynamoDB sloppy-quorum writes, Riak. Hints typically have a TTL (3 hours in Cassandra default) before anti-entropy takes over.
How Hinted Handoff Comes Up in Interviews
Scenario
DynamoDB-style cluster: one of the three replicas for a key is briefly down. You still want the write to succeed and the down node to catch up later.
A healthy peer accepts the write on the down node's behalf with a hint, then forwards it on recovery. Anti-entropy is the slower backstop; hinted handoff handles the common short-outage case quickly.
If the hint-holder also crashes before forwarding, Merkle-based anti-entropy is what guarantees eventual convergence.