Replicas distribution in the cluster


I wanted to understand how the replicas are distributed.

For e.g. in a cluster of 4 nodes(A,B,C,D), node A will be master of approximately one-fourth of the records. If I’ve configured a replication factor of 2, will the replicas of the records(whose master is node A) be distributed among all the remaining 3 nodes or only one of the other 3 nodes will be holding all the replicas?


The replicas are distributed across the remaining nodes. This minimizes the amount of data that is unavailable in a RF node outage, but has the disadvantage of ensuring some unavailability for any RF node outage.


This topic was automatically closed 6 days after the last reply. New replies are no longer allowed.