Centralized Approach of Load Balancing in Homogenous Grid Computing Environment

Abstract

There are two major challenges in Grid computing environment; managing the workload and resource management. To better facilitate to the Grid users it encompasses the other challenging areas like number of computing elements, nature of resources, autonomous nature of computing resources, processing capacities of the computing elements, various load conditions, overloading of jobs, underutilized of resources, etc. Increase the throughput and reduce the communication cost during job migration are major aspects of load balancing in grid computing environment. To balance the load we need an efficient and effective load-balancing algorithm that evenly distribute the load among the computing nodes. In this paper, a comparative analysis between the existing load balancing algorithms has been presented and compared with client server load balancing algorithm. The proposed algorithm has the monitored node to get the workload information from other connected computing elements nodes in the system and forward the information to the resource allocation node followed by load balancing. The Proposed algorithm reduces the complexity on transferring the jobs and significantly reduced the complexity, which is best suited for local communication within a local area network.