Depending on the volume configuration, data is distributed across multiple nodes in the cluster, removing a single point of failure, as no head or master node controls the cluster.
