基于改进RM算法的云环境任务调度研究  

Study on Cloud Computing Task Scheduling Strategy Based on RM Algorithm

作　　者：王祺元[1] 闫宏印[1] 

机构地区：[1]太原理工大学计算机科学与技术学院,太原030024

出　　处：《计算机测量与控制》2013年第6期1612-1614,共3页Computer Measurement &Control

基　　金：山西省自然科学基金(2010011025-2)

摘　　要：云计算中存在大规模的用户群和海量数据处理,如何高效地进行作业调度是云计算下需要解决的重要问题;通过对云计算下的Map-Reduce编程模型进行研究,着重分析了Map-Reduce作业调度算法,改进了一种单调速率计算能力调度算法;该算法结合作业周期和重要程度两个参数Ii和Ti来计算作业的优先级,并按照优先级分类作业调度;实验结果表明,此算法提高了作业调度的效率,是云计算中一种有效的作业调度算法。There are a large number of user groups and data processing in cloud computing, so how to conduct job scheduling effectively becomes the primary problem. During the research on Map--Reduce programming model, its job scheduling algorithm is emphatically ana- lyzed and a kind of improved rate--monotonic eomputing capability scheduling algorithm is proposed. This algorithm calculates the job＇ s priority through combining two parameters, including Ii and Ti, job period and importance degree, then in accordance with it to classify job scheduling. The simulation show that the algorithm improves the efficiency of job seheduling, which is an effective job scheduling algorithm in cloud computing.

关 键 词：云计算 单调速率 作业调度 优先级 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]