Command Palette

Search for a command to run...

Hectal
Phase 10Advanced11 of 17 in Redis

Replication & High Availability

Asynchronous replication, partial and full resync, replica lag and stale reads, Sentinel failover, split brain, and how to reason about CAP for a real Redis deployment.

Replication gives Redis read scaling and a standby to fail over to. It's asynchronous, which is why a Redis failover can lose acknowledged writes and why reading from replicas can show old data.

This phase makes those behaviours precise, then builds automatic failover with Sentinel and analyses what happens during network partitions, the part of Redis HA that interviews dig into hardest.

0/4 · 0%
4 topics ~38 min 8 code blocks & diagrams
Start with the first topic
1
10.1

Primary–Replica Replication

Replicas copy the primary asynchronously: a full sync with an RDB snapshot first, then a continuous stream of write commands. Replication IDs, offsets and a backlog let reconnecting replicas catch up with a partial resync instead of copying everything again.

9 min 1 diagram 1 code practice

2
10.2

Replica Lag, Stale Reads, and Lost Acknowledged Writes

Because replication is asynchronous, a replica may not yet have a write the primary acknowledged, so replica reads can be stale and a failover can lose acknowledged writes. You manage this with read routing, WAIT, min-replicas-to-write, and designs that tolerate the gap.

10 min 1 diagram 1 code practice

3
10.3

Redis Sentinel: Automatic Failover

Sentinel processes monitor a primary and its replicas, agree that the primary is down (SDOWN then ODOWN by quorum), elect a leader among themselves by majority, promote the best replica, reconfigure the others, and tell clients the new primary's address.

10 min 1 diagram 2 code practice

4
10.4

CAP, Network Partitions, and Split Brain

Don't label Redis "AP" or "CP". Analyse the specific deployment: during a partition, an isolated primary may keep accepting writes that are discarded when it rejoins (split brain), and settings like min-replicas-to-write and client behaviour decide how much. Redis replication is not consensus.

9 min 1 diagram practice