Cluster recheck interval causes unnecessary high load on cluster node

Issue

Currently, Pacemaker's failure-timeout resource meta-attribute, ocf:pacemaker:remote reconnect_interval instance attribute, and time-based configuration rules are not guaranteed to be checked any more often than the value of the cluster-recheck-interval cluster property. This leads to users setting cluster-recheck-interval very low, sometimes as low as 1min, which causes unnecessary load on the system during normal operation. Unless a new cluster event (resource failure, configuration change, etc.) intervenes, the failure will not be cleaned for 10 minutes.