作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2009, Vol. 35 ›› Issue (9): 31-33,3. doi: 10.3969/j.issn.1000-3428.2009.09.011

• 软件技术与数据库 • 上一篇    下一篇

基于动态多处理节点的分布式系统任务调度

梁 根1,秦 勇1,2,郭小雪3,梁活民4   

  1. (1. 茂名学院教育信息技术中心,茂名 525000;2. 华南理工大学计算机科学与工程学院,广州 510641;
    3. 茂名学院理学院,茂名 525000;4. 华南师范大学计算机学院,广州 510631)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2009-05-05 发布日期:2009-05-05

Task Scheduling in Distributed System Based on Dynamic Multi Processing Nodes

LIANG Gen1, QIN Yong1,2, GUO Xiao-xue3, LIANG Huo-min4   


  1. (1. Education Information and Technology Center, Maoming University, Maoming 525000; 2. School of Computer Science & Engineering, South China University of Technology, Guangzhou 510641; 3. College of Science, Maoming University, Maoming 525000; 4. College of Computer Science, South China Normal University, Guangzhou 510631)
    【Ab
  • Received:1900-01-01 Revised:1900-01-01 Online:2009-05-05 Published:2009-05-05

摘要: 针对固定处理节点分布式系统动态调控能力弱的问题,给出一种分布式系统任务调度模型,讨论单处理节点任务调度问题,提出平均处理强度指标,用于更准确地刻画处理节点的承载能力。推导出动态多处理节点的任务分配方法,优化分布式系统中任务处理的时间响应特性。模拟实验证明,该算法有较好的动态调控能力,能根据需要降低处理器负载、改善任务处理延时并更合理地利用系统资源。

关键词: 任务调度, 动态分配, 处理强度, 负载均衡

Abstract: Aiming at the problem of poor control ability of distributed system with fixed nodes, this paper gives a task scheduling model of distributed system. The problem of task scheduling in single node is discussed. In order to depict the capability of load-tolerance clearly, the processing intensity concept is drawn. The method of task scheduling among dynamic multi processing nodes is deduced to optimize the characteristic of responsibility time in task scheduling. Simulation experiments show that this algorithm has a better control ability, and it can reduce load of processor, improve the performance of the task-processed delay and use the resource of system more legitimately.

Key words: task scheduling, dynamic distribution, processing intensity, load balancing

中图分类号: