Resubmission Based Fault Tolerance Approach to Schedule Jobs in GRID Environment

Abstract

Grids are the type of distributed machines in which aggregation of resources is done to provide services. Grid environment needs to be fault tolerance so as to obtain maximum performance of the tasks that are being processed on it. A novel fault tolerance approach is designed by us. Fault tolerance approaches mainly lie into two categories, Replica Based approach, and Check-Pointing approach. The main limitation of Replica Based Approach (RBA) is its non-applicability to cost-based resources whereas Check-Pointing Approach (CPA) suffers from the inherent disadvantages of taking Check-Point which incur overhead and wastage of time. In our approach, the large task is divided into various subtasks on the basis of data flow and control flow dependencies. The experimental results show that the proposed approach is efficient than a CheckPoint approach in terms of various parameters like the number of Grid lets successfully completed an average execution time of a task. GridSim ToolKit 4.0 is used for simulation.