Abstract: Distributed systems play an important role on achieving good performance and high system utilization. The goal of a job scheduling system is to efficiently manage the distributed computing power of workstations, servers, and supercomputers in order to maximize job throughput and system utilization. There are many issues of distributed computing system which are discussed in this paper in brief. The main aim of the paper is to focus on fault tolerance and recover fault with less processing time. The proposed algorithm is assign tasks to other nodes only when master node moves from its original position using weights.

Keywords: Computing, master node time, e-cost, profit.