The grid computing main concern is to use the resources efficiently. For achieving this many grid resource scheduling algorithms are used for the efficient use of unused resources, especially CPU. The scheduling algorithms assign single complete job to a single resource. Instead, if these algorithms consider the degree of dependency among the modules of a job, they can be allocated parallel to the different resources. This reduces the completion time of the job and the resources can be utilized to its maximum extent. Towards this, the job scheduling using coupling algorithm is proposed. This algorithm puts forward the idea of considering the coupling degree while allocating the modules of a job parallel to different resources. In this algorithm, resource selection is done by using both its functional and non-functional properties. The algorithm works in 3 phases. It groups the interdependent modules of a job into different sets using coupling in the first phase. It checks the non-functional property i.e. availability of a resource using echo procedure in the second phase and in the third phase, the sets created in first phase, are allocated parallel to different available and matching resources. From the simulation results it is observed that job scheduling using coupling algorithm gives better performance in terms of reduced turnaround time as compared to First Come First Served, Largest Task First and Minimum Execution Time scheduling algorithms.
展开▼
