Memory Constrained Load Shared Minimum Execution Time Grid Task Scheduling Algorithm in a Heterogeneous Environment

Objective: To increase the resource utilization and balance the load in the grid environment. Methods: Memory Constrained Load Shared Minimum Execution Time (MCLSMET) scheduling is proposed to make best use of the resource utilization in a grid environment to reduce makespan. Load balancing is achieved by rescheduling the resources based on memory requirement and execution time of the tasks. This algorithm considers memory as Quality of Service (QoS) factor. Results: The proposed algorithm has been implemented in a simulated environment and the results are compared with the Minimum Execution Time (MET) algorithms. In MCLSMET algorithm, the Maximum Completion Time, Resource Utilization is computed to compare with the existing MET scheduling Algorithm. The MET scheduling algorithm produces the makespan 34 ms whereas the proposed method reduces the makespan to 15 ms for a task. In the existing MET scheduling Algorithm produces severe load imbalance problem. In the proposed method load is shared among the available resource and the resource utilization percentage is increased. Conclusion: The Memory Constrained Load Shared Minimum Execution Time (MCLSMET) scheduling algorithm is suggested that this algorithm produces higher resource utilization, reduces the makespan and load balancing.