A network partition refers to the failure of a network device that causes a network to partition. For example, in a network with multiple subnets were nodes A and B are located in one subnet and nodes C and D are in another. If the switch between the subnets fails, then the network is partitioned and nodes A and B can no longer communicate with nodes C and D.

The question is, what happens when a Cassandra cluster deployed across multiple subnets experiences a network partition? Cassandra’s fault tolerant mechanism allows the nodes in the partitioned subnets to continue to operate without disruption of service. Once the partitioned network is fixed and the nodes across the multiple subnets can communicate, Cassandra will resume replication and synchronize all the nodes in the cluster.