Grid Scheduling Based on Prediction of Task Completion Time  

作　　者：WU Yongwei CHEN Gang YANG Guangwen ZHENG Weimin 

机构地区：[1]Department of Computer Science and Technology, Tsinghua National Laboratory for Information Science and Technology, Tsinghua University, Beijing 100084, China

出　　处：《Chinese Journal of Electronics》2009年第3期389-394,共6页电子学报（英文版）

基　　金：This work is supported by the China Grid Project of Ministry of Education of China, the National Natural Science Foundation of China （No.60373004, No.60373005, No.90412006, No.90412011, No.60573110, No.90612016 and No.60673152）, National Key Basic Research Project of China （No.2004CB318000 and No.2003CB316907）, and National High Technology Development Program of China （No.2006AA01A101, No.2006AA01A108 and No.2006AA01Alll）.

摘　　要：Scheduling problems of grid research area are paid more and more attention recently. In this paper, a grid Scheduling model based on prediction of task completion time （SPCT） is proposed. Through Using Least Squares Discrete Curve Fitting, SPCT dynamically establishes the regression function of Completion time of task （CTT） according to the historical record first. Predicted completion time of each coming task is calculated for each candidate node with the regression function secondly. And then, the node with the least value will be allocated to run the task. The SPCT is used to input data sensitive applications and implemented in one real-world grid environment, Bioinformatics Grid Platform. Experimental result shows that the SPCT could reduce the average CTT of tasks by 19%.

关 键 词：Grid computing Job scheduling Task performance prediction Task modeling. 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] TP391