Pre-emptive resource allocation in Grid Computing (PRAG)

Computational grids have the potential for solving large scale scientific problems using heterogeneous and geographically distributed resources. Before this potential is realized, there are several technical overheads that must be overcome. This paper introduces PRAG - a resource allocation technique that is pre-emptive in nature. The critical problem for grid users is efficient allocation of resources to jobs that is effective utilization of computational grids and a certain Quality of service (Qos) for grid users. This paper addresses those problems by a pre-emptive technique for resource allocation using advance reservation planning and time limit (duration of a job after that its priority is incremented). The grid agent performs job scheduling based on the Advance reservation request, and immediate reservation request on allocated resources. The main aim of this technique is to minimize the total time to execute all jobs, minimize the waiting Time by using pre-emptive scheduling that eliminates starving of jobs in Global Queue, maximize resource utilization rate and balance the load among resources.