通信学报 ›› 2012, Vol. 33 ›› Issue (Z1): 165-169.doi: 10.3969/j.issn.1000-436x.2012.z1.021

• 学术论文 • 上一篇    下一篇

基于S4框架的并行复杂事件处理系统

陈皓1,2,李瑜1,2,虎嵩林1,梁英1   

  1. 1 中国科学院 计算技术研究所,北京 100190
    2 中国科学院 研究生院,北京 100049
  • 出版日期:2012-09-25 发布日期:2017-08-03
  • 基金资助:
    国网信息通信有限公司科技基金资助项目;广东省科技计划基金资助项目

Parallel complex event processing system based on S4 framework

Hao CHEN1,2,Yu LI1,2,Song-lin HU1,Ying LIANG1   

  1. 1 Institute of Computing Technology,Chinese Academy of Sciences,Beijing 100190,China
    2 Graduate University,Chinese Academy of Sciences,Beijing 100049,China
  • Online:2012-09-25 Published:2017-08-03
  • Supported by:
    Science Plan of State Grid Information and Telecommunacation Co,Ltd;The Planned Science and Technology Project of Guangdong Province

摘要:

针对复杂事件处理技术在单机的吞吐量瓶颈,而现有通用并行框架不适合复杂事件处理系统的问题,在分析现有并行技术的基础上,设计了14种复杂事件处理操作符,提出了以操作符为单位并行的复杂事件处理系统框架,给出了基于操作符的事件流负载分流方法。并基于S4系统实现了并行复杂事件处理系统,提供比单机运行更高的可靠性和吞吐率。通过集群上的实验证明,在一定范围内,该系统的吞吐量可随集群节点数增加而线性增长。

关键词: 复杂事件处理, 海量数据, 并行计算

Abstract:

There was a throughput bottleneck for complex event processing in only one computer,and the existing general-purpose parallel framework was not suitable for complex event processing system.Based on the analysis of existing parallel technology,fourteen kinds of complex event processing operator and a parallel framework by operator for complex event processing system were proposed,the event stream load triage based on operator was given.S4 system for parallel complex event processing system was utilized,provided greater reliability and throughput than stand-alone operation.Through the cluster experiments show that,within a certain range,the throughput of the system can be increased with the number of cluster nodes linear growth.

Key words: complex event processing, mass data, parallel computing

No Suggested Reading articles found!