Question: What are the types of replication in Redis?

Answer

Redis supports several replication models to enhance performance, data redundancy, and high availability. Below are the types of replication available in Redis:

Master-Slave Replication: The most basic form of replication in Redis is master-slave replication. Data from the master server is replicated to one or more slave servers. Slaves can serve read operations, but all write operations are performed on the master.

redis-server --port 6379
redis-server --port 6380 --slaveof localhost 6379

Sentinel System: While not a direct replication type, Redis Sentinel provides a system for managing Redis instances. It monitors master and slave instances, providing automatic failover if the master goes down. Sentinels can elect a new master from among the slaves and reconfigure other slaves to use the new master.

Cluster Mode: Redis Cluster provides a way to run a Redis installation where data is automatically sharded across multiple Redis nodes. It allows for the continuation of operations even when some nodes fail or are unreachable. Redis Cluster also supports replication by having each shard (master) be able to have one or more replica nodes (slaves).

Diskless Replication: Starting with Redis 2.8.18, diskless replication became an option. With this configuration, the data is streamed directly from the master's memory to the slave's memory without being written to the disk on the master side first. This can provide faster synchronization times.

# On master
replica-serve-stale-data no
repl-diskless-sync yes
repl-diskless-sync-delay 5

# On slave
replica-priority 100

Partial Replication: With partial replication, if a connection between master and slave breaks temporarily, upon reconnection, only the part of the data that was missed during the disconnection will be transmitted to the slave, assuming the master still has the relevant portion of the dataset in its buffer.

Each replication type serves different use-cases and operational needs, such as load balancing, failover strategies, data persistence, or sharding setups.