Under normal conditions all peers participate equally in sharing the workload.

Under peer failure conditions all remaining peers participate equally in sharing the extra workload.

When healing failed peers that have rejoined the cluster, the entire cluster shares the workload.