A shared heterogeneous computing is comprised in distributed system & data resources are distributed across the network boundaries. Owing to geographically distributed resources, which is heterogeneous in nature, owned by different individuals or organizational that owns policies with varying loads & the Resource Management system caused by the availability, even obtaining the QOS in the resource faults existence a comple which overhead of checkpoint by clustering based checkpoint and optimize the number of nodes in cluster by Flower pollination algorithm and use centrality base trust case depend on connectivity and importance of node using game theory approaches.