Science and Development Network (SciDev.Net)
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based on selecting resources that have less resource response time or less grid bandwidth while these resources may be out of date at the scheduling time.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.