计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (20): 265-270.DOI: 10.3778/j.issn.1002-8331.1708-0044

• 工程与应用 • 上一篇    

面向司法专题分析的大数据计算资源调度方法

刘忠麟1,2,艾中良1,2,高  泽1,李常宝1   

  1. 1.华北计算技术研究所,北京 100083
    2.中国司法大数据研究院有限公司,北京 100083
  • 出版日期:2017-10-15 发布日期:2017-10-31

Large data computing resource scheduling method for judicial analysis

LIU Zhonglin1,2, AI Zhongliang1,2, GAO Ze1, LI Changbao1   

  1. 1.North China Institute of Computing Technology, Beijing 100083, China
    2.China Justice Big Data Institute CO., Ltd, Beijing 100083, China
  • Online:2017-10-15 Published:2017-10-31

摘要: 针对司法专题分析过程中面临的交互式分析类数据处理执行效率低的问题,提出了一种基于任务类型的计算资源调度方法,为任务类型建立计算资源配额管理机制。在类型配额内具备抢占式优先调度权,在类型配额外可以借用其他任务类型的空闲资源。实验与分析表明,该方法能够在兼顾普通大数据处理任务执行效率的前提下显著提升交互式分析类任务的执行效率。

关键词: 计算资源调度, 任务类型, 交互式分析

Abstract: Aiming at the problem of low efficiency in interactive data processing of judicial specialties analysis, a computing resource scheduling method based on the job type is proposed. The method establishes the quota management mechanism for the job type. Within its quota range the job has the preemptive priority scheduling right, beyond its quota range the job can borrow the idle computing resource of other types. Experiments and analysis show that this method can significantly improve the efficiency of interactive analysis jobs taking into account the efficiency of common data processing jobs.

Key words: computing resource scheduling, job type, interactive analysis