华南理工大学学报(自然科学版) ›› 2010, Vol. 38 ›› Issue (11): 122-127,134.doi: 10.3969/j.issn.1000-565X.2010.11.022

• 计算机科学与技术 • 上一篇    下一篇

一种可用性增强的计算网格作业执行模型

胡周君 胡志刚   

  1. 中南大学 信息科学与工程学院, 湖南 长沙 410083
  • 收稿日期:2009-11-02 修回日期:2010-04-14 出版日期:2010-11-25 发布日期:2010-11-25
  • 通信作者: 胡周君(1982-),男,博士生,主要从事网格计算研究. E-mail:sunddy.h@163.com
  • 作者简介:胡周君(1982-),男,博士生,主要从事网格计算研究.
  • 基金资助:

    国家自然科学基金资助项目(60970038;60673165)

Availability-Enhanced Model for Job Execution in Computational Grids

Hu Zhou-jun  Hu Zhi-gang   

  1. School of Information Science and Engineering,Central South University,Changsha 410083,Hunan,China
  • Received:2009-11-02 Revised:2010-04-14 Online:2010-11-25 Published:2010-11-25
  • Contact: 胡周君(1982-),男,博士生,主要从事网格计算研究. E-mail:sunddy.h@163.com
  • About author:胡周君(1982-),男,博士生,主要从事网格计算研究.
  • Supported by:

    国家自然科学基金资助项目(60970038;60673165)

摘要: 为增强服务网格中的用户QoS水平,文中针对由大量可用性不高的资源组成的网格环境,提出了可用性增强的作业执行服务(AJES)的概念及确定方法,基于该概念提出了可用性增强的作业执行模型(Availability-enhanced Model for Job Execution,AMJE),通过将AJES中多个网格节点的资源分配给作业冗余执行,以增强其可用性.在AMJE中,不同参与方根据各自利益进行博弈,求得了折中的AJES可用性阈值.实验结果表明,文中提出的作业服务模型能显著提高资源收益,增强用户QoS的保障水平,且对于可用性动态变化的网格环境具有更好的适应性.

关键词: 计算网格, 可用性, 服务质量, 作业执行模型

Abstract:

In order to improve the quality of service(QoS) in service grid environment,the concept of Availability-Enhanced Job Execution Service(AJES) and its constructing method are proposed for the grid environment consisting of a large amount of low-availability resources.Then,an availability-enhanced model for job execution(AMJE) is derived from AJES,which redundantly schedules the resources to one or more grid sites within AJES,thus enhancing the availability required for job execution.In AMJE,different participants on behalf of their profits are described as a game,and a balance threshold of AJES availability is obtained.Experimental results show that AMJE increases the revenue of resource providers,improves the QoS level for end users,and exhibits good adaptability to dynamic grid environment.

Key words: computational grid, availability, quality of service, job execution model