Effective Scheduling of Jobs Using Reallocation of Resources Along With Best Fit Strategy and Priority

Abstract

Grid computing let the devices to distribute the job across different environment, where the resources are heterogeneous in nature. It is a collection of resources from different states in order to attain common goal. Scheduling of jobs is one of the major crises that play a vital role nowadays. So, this paper proposes a scheduling algorithm where the resources can be effectively allocated. Grid Scheduler plays a vital role in allocating the jobs to the resources. Before the job arrives, the function of the Grid Scheduler is to reassemble the memory by its size. When the job arrives, the Shortest Job First algorithm along with the best fit strategy is applied to allocate the job effectively to the resources in order to reduce the wastage of memory. If same sized jobs arrive at the scheduler priority rule is applied to allocate the job. If the job size is not as same as the resource size, at that case the Grid Scheduler locates the resources that are nearby equal to the job size and allocate the resources for that job. The free space that is remaining in those resources is likewise combined with all the free spaces that are remaining in all of the resources, so that these spaces can be combined and used as resource for some other job allocation. The memory wastage is managed at this situation. Here Grid Resource Broker is used for combining all the free memory space.
Through this approach scheduling can be performed effectively.