摘要
针对文中描述的计算网格资源环境模型,构造了一种分布式的层次型任务调度模型,任务调度分为计算资源站点的选择以及资源站点内部的本地调度两层进行。通过研究该调度模型,提出了一种基于双目标衡量函数的资源选择算法,该算法可以通过设置相关参数动态调节响应时间和价格在总目标中所占比重。试验结果表明能够选择综合满足响应时间和价格这两个目标的计算资源,以适应用户的不同需求。
Aiming at the computational grid resource environment model described in this paper,a kind of distributed layered task scheduling model is constructed,which divides task scheduling into two layers：selection of computational resource site and local scheduling inside a resource slte.By researching this scheduling model,a computational resource selection algorithm based on double goals measurement function is proposed,which can dynamically adjust both response time and price weights in the total goal through setting related parameters.Experiment results indicate that this algorithm is able to pick up the computational resources which synthetically meet response time and price goals ,so as to adapt to different user requirements.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第34期117-119,206,共4页
Computer Engineering and Applications
基金
国家自然科学基金重点项目(编号:60433020)
中南大学博士后资助项目
关键词
计算网格
任务调度
响应时间
价格
资源选择算法
computational grid,task scheduling,response time,price,resource selection algorithm
