文章基本信息

标题：OPT-Min-Min Scheduling Algorithm of Grid Resources
本地全文：下载
作者：Cao, Lijun ; Liu, Xiyin ; Wang, Haiming 等
期刊名称：Journal of Software
印刷版ISSN：1796-217X
出版年度：2014
卷号：9
期号：7
页码：1868-1875
DOI：10.4304/jsw.9.7.1868-1875
语种：English
出版社：Academy Publisher
摘要：As a newly emerged distributed computing platform, the Gird aims at the implementation of resources sharing and collaborative computing on the internet. Research on grid resource scheduling algorithm is one of the core problems of grid technology, as well as an essential part of grid resource management. Resource scheduling in grid environment refers to the application of tasks scheduling algorithm to assign the works in the grid to the right resources. During the allocation process, the computing performance of grid resource nodes, communication parameters and loading balance, etc., shall be put into consideration. Since they are all dynamic for both the grid resources and the grid resource nodes, requirements on grid resource scheduling algorithm are relatively higher if compared to those existing parallel resource scheduling algorithm. Based on the Min-Min scheduling algorithm, the OPT-Min-Min scheduling algorithm is presented to overcome the drawbacks. Upon the basis of applying Min-Min pre-scheduling into stage one, by adapting the strategy of two-rounds scheduling, the assignment on heavy load resources are rescheduled to balance the load. Actual cases are used to illustrate the superiority of OPT-Min-Min scheduling algorithm to the Min-Min scheduling algorithm. Simulation experiments were carried out to the batch grid resource scheduling algorithms including Min-Min, Max-Min, Min-mean and OPT-Min-Min. According to the ETC generation technique used in experimental benchmarks, ETC matrix was generated. By the comparing and analyzing the Min-Min, Max-Min and Min-mean scheduling algorithm, the validity of OPT-Min-Min scheduling algorithm is proved.
关键词：Grid;Scheduling Strategy;OPT-Min-Min;resource scheduling algorithm;load balance