In case of sequential node failure or shutdown, this allows the cluster to continue running with less than a majority of ...