Scheduling Jobs in Face of Status Update Timing of Resources in Computational Grids

Users and resources frequently join and leave computational grid, hence the state of the grid changes dynamically. So, an effective job scheduling strategy is needed that consider the dynamically changed conditions in the grid. Most of the existing scheduling strategies are mainly based on selecting resources that have less resource response time or less grid bandwidth while these resources may be out of date at the scheduling time. This paper proposes a scheduling strategy with the objective of minimizing the effects of grid dynamism on the grid performance. The strategy schedules jobs to the resources according to status update timing combined with response time of the resources. The proposed strategy is compared with the time-based strategy in terms of the number of jobs executed successfully within the specified deadline and the makespan of user applications. Experimental results have shown that the proposed strategy can considerably improve the performance of the grid.