Efficiently utilization of redundancy backup server by forming dynamic clustering in Distributed Systems for tolerating faults

This paper proposes a novel REDENDENCY THROUGH BACKUP PROCESS with CLUSTERING scheme which aims to address the following concerns: reliable, effective with high maintenance backup can be achieved by constructing a multi-node clustering backbone with a small number of backup cluster-heads for redundancy filling system through backup process for FAULT TOLRANCE DISTRIBUTED SYSTEMS. We can successfully reduce the overhead of backup servers and enhance the speed of backup delivering in an allowable time span compared to other redundancy fault tolerance distributed operating systems for both WAN and LAN networks. Next, we utilize the CLUSTER STRUCHING MECHANISM which determines the cluster size according to the leaving frequency of cluster-members. As the number of leaving events is reduced, the cluster topology is more stable and the BACKUP may also be available in a short and manageable time span before all the distributed system becomes failed. We have done our simulation using MATLAB to show cluster formation with a back-up server election and have investigated the performance of our system for different scenarios.