Abstract:
The problem of unbalanced of metatask scheduling in the performance predicted and scheduling system, Grid Harvest Service(GHS), are researched. In order to solve the problem, a Heuristic Min-Min and Max-Min algorithm(H-MM) is proposed and implemented into GHS to schedule tasks. Experimental results show that the algorithm overweighs traditional algorithms a lot in makespan and scalability parameters.
Key words:
grid,
task scheduling,
grid harvest service
摘要: 针对非专用网格性能预测调度系统(GHS)中元任务调度算法负载不平衡问题,提出一种启发式动态任务调度策略(H-MM),结合Min-Min与Max-Min的优点,将H-MM植入GHS,结果表明H-MM在时间跨度与可扩展性方面较以往算法有较大提高。
关键词:
网格,
任务调度,
网格丰收服务
CLC Number:
Wushou Sliamu; HOU Yong; ZHOU Yan-hui. Improvement of Task Scheduling Algorithm Based on GHS[J]. Computer Engineering, 2009, 35(9): 34-36.
吾守尔.斯拉木;侯 勇;周艳慧. 基于GHS的任务调度算法改进[J]. 计算机工程, 2009, 35(9): 34-36.