A two-level hierarchical scheduling method for independent tasks in grids

Parallel and distributed computing environments are essential and utilized to meet the needs of a wide variety of high-throughput applications. Scheduling strategies are important in order to efficiently utilize resources and to improve response times, throughput and utilization of computing platforms. In this paper, we present a two-level hierarchical method for scheduling of independent coarse-grained tasks in grid environments. With two-level architecture, the Grid Scheduler (master node of upper-level) proceeds with distribution of tasks to computing sites, while the Local Resource Manager (master node of lower level) assigns this task to an available computing node according to a given threshold. Comparing experimental results with those obtained from well known traditional scheduling algorithms, the effectiveness of the proposed method consistently shows a benefit from this approach.