Authors : L.M. Nithya and A. Shanmugam
Abstract: In this study, it is proposed to enhance the grid job scheduling algorithm by including economic parameters-deadline and budget that schedule the job to a suitable resource which is capable of satisfying not only computational requirements but also economic constraints of a job. As grid environment is heterogeneous in nature, there may be wide range of failures that might affect job execution. In order to handle failures and to avoid rescheduling of jobs after the detection of failure, pro-active failure handling methods are included in the enhanced scheduling algorithm to estimate the availability of resources in the grid and also to preemptively calculate the expected long term capacity of the grid. The performance indicators such as average waiting time and queue completion time are calculated for the proposed job scheduling algorithm and backfill algorithm, comparison exemplifies that the proposed algorithm gives significant results. The proposed scheduling algorithm with failure handling methods outperforms the scheduling algorithm without failure handing in most of the situations. It has increased the job processing rate by 58%, decreased the job failure rate and job rejection rate by 66 and 16%, respectively.
L.M. Nithya and A. Shanmugam, 2011. Job Scheduling with Economic Parameters and Failure Handling Methods in a Grid Computing Environment. Research Journal of Applied Sciences, 6: 525-531.