US7296191B2 - Scalable Method of Continuous Monitoring the Remotely Accessible Resources Against the Node Failures for Very Large Clusters