Replicating the Data in Key-Value Store [merged with scalability]

Replication

We have various methods to replicate the storage. It can be either a primary-secondary relationship or a peer-to-peer.

In primary-secondary, one of the storage is primary and other storage(s) is/are secondary, and secondary replicates its data from primary. The primary serves the write requests while the secondary serves read requests. After writing, there’s a lag for replication. Moreover, if the primary goes down, we cannot write into the storage and it becomes a single point of failure.