Abstract

Grid computing is a conventional technology to integrate large scale distributed resource sharing. It harnesses the idle time of resources for effective Utilization. Task scheduling is an essential problem for achieving high performance in grid computing systems. Many parameters like execution time, memory capacity, network bandwidth, communication cost etc affect the task scheduling in Grid environment. An effective and efficient scheduling algorithm is needed to attain the promising potentials of significant distributed resources and to improve the overall throughput of the Grid environment. Minimum Execution Time scheduling algorithm fails to produce a load balanced schedule as well as it considers only the execution time of tasks for scheduling. This work considers the Quality of Service (QoS) factor such as memory requirement of task and the execution time of the task to make the algorithm efficient in real time environment. This Dual Objective Load shared Minimum Execution Time [DOLSMET] produces an efficient schedule that balances the load as well as satisfies the memory requirements of tasks