High availability

In addition to the high scalability enabled by the CRUSH algorithm, Ceph must also maintain high availability.

Maintaining high availability means that Ceph clients must be able to read and write data even when the cluster is in a degraded state, or when a monitor fails.