Authors : Ravin Ahuja , G. Gabrani and Asok De
Abstract: In the grid technologies, it has become possible to allow many users or applications having multiple jobs to seamlessly access and share huge and heterogeneous pools of computational data, storage and network resources that are geographically distributed. This research is into the domain of problems arising out of making resources available not only time specific but also to cater to the needs of specific users with different priorities at a particular time. In the present dynamic scenario when jobs (belonging to different users with different priorities) are aplenty, the situation arising confronts a challenge regarding scheduling of the incoming jobs after resolving their priorities and allocating them to the resources. These jobs can be allocated to the resources either immediately (current reservation) or for some time in future (advance reservation). This study presents, a Dynamic Priority Scheduler for Advance Reservation (DPSAR) in a space shared environment to coordinate the resource sharing in distributed grid computing environments. The DPSAR aims at analyzing the multifarious jobs and sequel to calibrating on the basis of priorities and protocols further reserving them to the respective resources. In this study, we have simulated the DPSAR and its results are analyzed to measure the performance issues. The results confirmed to an improved performance in terms of number of rejections and resource utilization over an existing scheduler based on advance reservation technique (without priority) and pit falls were found to be nominal.
Ravin Ahuja , G. Gabrani and Asok De , 2009. A Dynamic Priority Scheduler for Advance Reservation in Grid Computing. International Journal of Soft Computing, 4: 60-67.