西安电子科技大学学报

• 研究论文 • 上一篇    下一篇

混合关键任务可靠调度方法与调度性分析

景维鹏1,2;霍帅起1,2;陈广胜1,2;刘亚秋1,2   

  1. (1. 东北林业大学 信息与计算机工程学院,黑龙江 哈尔滨 150040;
    2. 黑龙江省林业生态大数据存储与高性能(云)计算工程研究中心,黑龙江 哈尔滨 150040)
  • 收稿日期:2016-03-16 出版日期:2016-12-20 发布日期:2017-01-19
  • 作者简介:景维鹏(1979-),男,副教授,博士,E-mail: nefujwp@163.com.
  • 基金资助:

    中央高校基本科研业务费专项资金资助项目(2572014EB05-4);黑龙江省自然科学基金重点资助项目(ZD201403);林业公益性行业科研专项经费资助项目(201504307)

Novel mixed-criticality reliability scheduling strategy and schedulability test

JING Weipeng1,2;HUO Shuaiqi1,2;CHEN Guangsheng1,2;LIU Yaqiu1,2   

  1. (1. The College of Information and Computer Engineering, Northeast Forestry Univ., Harbin 150040, China;
    2. Heilongjiang Province Engineering Technology Research Center For Forestry Ecological Big Data Storage and High Performance (Cloud) Computing, Harbin 150040, China)
  • Received:2016-03-16 Online:2016-12-20 Published:2017-01-19

摘要:

为了解决云计算环境下混合关键性任务的可靠调度问题,提出了一种基于主副版本两阶段的混合关键任务可靠调度方法.算法首先对需要调度的混合关键性任务进行优先级划分,按照调度截止期最短的原则将主版本任务调度到目标虚拟机上,对副版本任务按照复制成本最低的原则使用重叠方法进行调度;再对调度到不同虚拟机上的主副版本任务进行可调度分析,对于不能满足分析的任务启动更高关键性等级进行处理.实验结果表明了混合关键任务可靠调度方法具有较高的可靠性和负载平衡能力.

关键词: 云计算, 混合关键性任务, 可靠调度, 主副版本

Abstract:

In order to solve the reliable scientific workflow scheduling problem for the Mixed-Criticality task in cloud computing, we proposed the Mixed-Criticality reliability scheduling strategy (MCRSS) based on Primary/Backup. First, the priority of the primary Mixed-Criticality task is determined and the task is scheduled for the virtual processor with the deadline being the shortest, the backup is the virtual processor with the cost of copy being the lowest. Second, the schedulability test of the primary and backup task are proposed. If the task does not satisfy the schedulability test, then the task will change to high criticality. Experimental results show that the MCRSS algorithm is of high reliability and load balancing capabilities.

Key words: cloud computing, mix-criticality task, reliable scheduling, primary/backup