A Compendium of Heuristic Methods for Scheduling in Computational Grids

摘要

Scheduling in large scale distributed computing environments such as Computational Grids, is currently receiving a considerable attention of researchers. Despite that scheduling in such systems has much in common with scheduling in traditional distributing systems, the new characteristics of Grid systems make the problem more complex and versatile to match different needs of Grid-enabled applications. In this work, by conceiving scheduling problem as a family of problems, we first identify most common versions of the scheduling problem based on six dimensions: type of the environment, architecture type of the scheduler, immediacy of the processing, type of interrelations among tasks, type of preemptive policy and type of optimization model. Then, we review different families of heuristic methods used for the resolution of the problem, including ad hoc methods, local search methods and population-based methods.