一种优先满足最小服务需求的动态网格资源调度算法

A Dynamic Resource Scheduling Algorithm Based on the Minimum Service Requirement

资源调度问题是网格研究必须解决的关键问题之一。目前,围绕着网格中的资源调度算法,国内外已做了大量的研究工作,先后提出了各种静态和动态调度算法。本文针对目前网格调度机制存在的问题,介绍了一种新的网格调度技术——优先满足最小服务需求的动态网格资源调度算法。该调度算法优先满足现有任务对资源的最小要求,从而减小单个任务的等待时间。实验结果表明,该方法不但可以有效减少单个任务的延迟,而且在任务的吞吐率及CPU效率方面都比较好。

Resource scheduling is a critical problem involved in grid studies. Currently, much research can be found about grid scheduling, and many scheduling algorithms including static and dynamic scheduling algorithms have been proposed. In this paper, a scheduling algorithm based on the minimum service requirement is presented to capture the dynamics and impact of simultaneously co-allocated tasks in a grid. In this approach, the grid scheduler first archives every task＇s minimum service requirement, which decreases each task＇s waiting time. The experimental results demonstrate that this approach diminishes each task＇s latency and contributes to the overall grid task throughput and CPU efficiency.