泛集群环境中计算密集型任务流调度策略
CSTR:
作者:
作者单位:

(1. 东北石油大学计算机信息与技术学院,黑龙江大庆163318;2. 黑龙江省石油大数据与智能分析重点实验室,黑龙江大庆163318)

作者简介:

通讯作者:

E-mail: csli_dmis@nepu.edu.cn.

中图分类号:

TP13

基金项目:

国家自然科学基金面上项目(51774090, 61502094);黑龙江省教育厅科研专项创新基金项目(2017YDL-12);东北石油大学青年科学基金项目(2017PYQZL-11).


Scheduling strategy of compute-intensive task-flow in generalized cluster
Author:
Affiliation:

(1. College of Computer & Information Technology,Northeast Petroleum University,Daqing163318,China;2. Heilongjiang Provincial Key Laboratory of Oil Big Data & Intelligent Analisys,Daqing 163318,China)

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对计算节点较多的泛集群环境下难以快速、合理地制定计算密集型任务流调度方案的问题,提出一种基于多目标连续竞买博弈的任务调度策略.建立多目标优化调度模型,降低多目标优化函数维度,并采用线性加权和法将其转化为总和目标函数,以保证最优解的合理性.为提高最优解搜索速度,引入ETC矩阵作为最优解表达形式,设计连续竞买博弈算法.模拟真实场景并通过与同类算法的对比,表明了调度策略在泛集群环境下的响应速度、资源性价比和总成本支出等方面具有明显优势.

    Abstract:

    A scheduling strategy based on multi-objective continuous bidding game is proposed to solve the problem that it is difficult to make a quick and reasonable scheduling plan for compute-intensive task-flow in a generalized-cluster with many computing nodes. To ensure the rationality of the optimal solution, a multi-objective optimal scheduling model is established, the dimensions of multi-objective optimization function are reduced, and the multi-objective optimization function is converted into a sum-objective function using the linear weighting method. For improving the search speed of the optimal solution, the ETC matrix is introduced for expressing the form of optimal solution, and continuous bidding game algorithm is designed. By simulating real scenarios and comparing with similar algorithms, it is proved that the scheduling strategy has obvious advantages regarding response speed, resource cost performance and total cost expenditure in the generalized-cluster.

    参考文献
    相似文献
    引证文献
引用本文

张可佳,胡亚楠,李春生,等.泛集群环境中计算密集型任务流调度策略[J].控制与决策,2019,34(12):2537-2546

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2019-12-04
  • 出版日期:
文章二维码