Settings in a failover cluster that determine the point at which too many failures will stop the cluster from running.