QoS based Task Rescheduling in Computational Grid Environment

Grid computing is an emerging field for the next generation of parallel distributed computing platform for solving large scale computational and data intensive problems. In general, the grid users need not be aware of the computational resources that are used for executing their applications and storing their data. In grid environment, all the tasks are not possible to run in all the resources. A challenging task in grid computing is to ensure the guaranteed Quality of Service (QoS) in grid scheduling. Thus the implementations of QoS based algorithms are necessary. The main objective of grid scheduling is to find a feasible schedule that minimizes the makespan. In recent times, QoS based heuristic grid scheduling algorithms have become a vibrant research topic. So, in this paper, we proposed a QoS based Task Rescheduling algorithm (QTR), in which the scheduled but not executed tasks are collected from the heuristic algorithms and rescheduled them using Minimum Completion Time (MCT) algorithm. The simulation model with twelve benchmark instances has been used for testing the performance of the proposed algorithm. The results of the computations show that the QTR algorithm can produce solutions with reduced makespan compared to the existing approaches.