• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2014, Vol. 36 ›› Issue (12): 2272-2279.

• 论文 • 上一篇    下一篇

胖树的网络演算模型

秦广军,祝明发,肖利民,阮利   

  1. (1.北京航空航天大学软件开发环境国家重点实验室,北京 100191;2.北京航空航天大学计算机学院,北京 100191)
  • 收稿日期:2014-08-30 修回日期:2014-11-01 出版日期:2014-12-25 发布日期:2014-12-25
  • 基金资助:

    北京市自然科学基金资助项目(4122042);国家教育部博士点专项基金资助项目(20101102110018);国家自然科学基金资助项目(61370059)

Network calculus for fattree network           

Qin Guangjun,ZHU Mingfa,XIAO Limin,RUAN Li   

  1.   (1.State Key Laboratory of Software Development Environment,Beihang University,Beijing 100191;
    2.School of Computer Science and Engineering,Beihang University,Beijing 100191,China)
  • Received:2014-08-30 Revised:2014-11-01 Online:2014-12-25 Published:2014-12-25

摘要:

网络演算近年来已成为网络领域重要的性能分析工具之一,但是,传统的网络演算主要关注网络的QoS性质,通过计算数据流的最大数据积压、最大端到端延迟、服务曲线等来分析网络的性能。在高性能计算系统中,则主要关心网络系统的吞吐率、通信延迟以及网络的饱和参数等,传统的网络演算理论无法有效计算这些参数。将网络演算与排队论结合,定义和推导了网络演算下的通信延迟和吞吐率,构造了一套适合高性能网络分析的网络演算分析方法。基于设计的方法,通过对胖树网络下均匀流量模式的分析,表明本文所设计的方法能够有效分析高性能网络的通信延迟、吞吐率等参数,并能有效捕捉网络的饱和状态,与模拟器的对比结果也表明本文模型能够基本吻合仿真结果。

关键词: 胖树网络, 网络演算, 通信延迟, 吞吐量, 流量模式

Abstract:

 Network calculus is one of the most important performance analysis tools in the network research area. The classical network calculus is mainly focused on the QoS, and the maximum backlog, the maximum network delay and the service curve are the significant analyzing metrics. However, the researchers in the highperformance computing systems are more concerned about the throughput, the communication latency, the network saturation point, etc., which cannot be inferred from the classical network calculus. The queuing theory is introduced into the network calculus, the communication latency and the throughput formulas are derived, and a network calculus analysis method for highperformance network calculus is proposed. Based on the analysis method, the uniform traffic pattern of the fattree network is analyzed. The analysis results indicate that the method can efficiently describe the network communication process and capture the saturation point, and basically be consistent with simulated results.

Key words: fat-tree network;network calculus;communication latency;throughput;traffic patterns