Durability for Rediss
Redis durability describes how data survives restarts and failures, through persistence on disk and replication to other nodes. RDB takes point-in-time snapshots, AOF logs every write, and replicas copy data from a primary. Redis Sentinel can detect a failed primary and promote a replica automatically.
Stage: Reliability · Part of the Redis roadmap · Last reviewed 19 November 2026
What to learn in Durability
- Compare RDB snapshots and AOF logging and the data each can lose
- Configure appendfsync and explain the trade-off between safety and speed
- Set up a replica using the replicaof setting
- Explain asynchronous replication and why a failover can lose recent writes
- Describe how Sentinel monitors and promotes a new primary
Durability interview questions
How do RDB and AOF persistence differ?
RDB writes compact snapshots at intervals, so restarts are fast but writes since the last snapshot can be lost. AOF appends each write command to a log and can fsync every second or every write, losing less data but producing larger files. Many deployments enable both.
Is Redis replication synchronous?
By default it is asynchronous. The primary acknowledges a write to the client without waiting for replicas, so if it fails before replicas copy the write, that write is lost after failover. The WAIT command can request acknowledgements from replicas but still does not make the system strongly consistent.
Where Durability fits in the Redis roadmap
Durability belongs to the Reliability stage (intermediate level), which takes about 9 hours. Persistence and scale.
Other units in this stage
Back to the full Redis roadmap