数据中心网络多路径路由算法

删除或更新信息，请邮件至freekaoyan#163.com(#换成@)

清华大学辅仁网/2017-07-07

数据中心网络多路径路由算法

杨洋^1,2, 杨家海¹, 秦董洪³

1. 清华大学网络科学与网络空间研究院, 北京 100084;
2. 西安通信学院, 西安 710106;
3. 广西民族大学信息科学与工程学院, 南宁 530006

Multipath routing algorithm for data center networks

YANG Yang^1,2, YANG Jiahai¹, QIN Donghong³

1. Institute for the Network Sciences and Cyberspace, Tsinghua University, Beijing 100084, China;
2. Xi'an Communication Institute, Xi'an 710106, China;
3. School of Information Science and Engineering, Guangxi University for Nationalities, Nanning 530006, China

摘要:

输出: BibTeX | EndNote (RIS)

摘要数据中心网络流量分布的不均衡增加了网络拥塞产生的可能性, 由于数据中心网络的流量特性, 使得传统IP网络的流量工程方法不一定适合。该文在SDN/OF(software defined network/OpenFlow)的结构下, 提出了一种基于多路径传输的动态路由算法 (dynamic routing algorithm based on multipath propagation, Dramp)并作为SDN/OF结构中应用层的流量均衡策略。该算法在重新定义链路关键度并求解链路权值优化问题的基础上, 能充分利用数据中心网络中存在的冗余路径, 在完成细粒度流量均衡的同时, 能很好地克服控制器的计算开销, 完成路由优化的目标。通过在Mininet仿真平台中部署并进行仿真实验, 与等开销多路径路由算法ECMP(equal-cost multi-path)以及GFF(global first fit)路由算法相比较, 结果展示了Dramp的优越性能, 同时证明了在数据中心网络中采用Dramp作为流量工程的解决方案更简单、更实用。

关键词 ：网络拥塞,软件定义网络(SDN),链路关键度,多路径路由,流量均衡

Abstract：Unbalanced distributions of traffic in the data center networks increase the network congestion. The special traffic characteristics in data center networks reduce the effectiveness of traditional IP network traffic engineering methods. This paper presents a dynamic routing algorithm based on multipath propagation (Dramp) using software defined for network/OpenFlow (SDN/OF) frameworks. This algorithm makes efficient use of multiple paths by redefining the degree of the critical link and solving an optimization problem using link weights. The objective is to reduce the overhead in the control plane for a fine-grained traffic balance for routing optimization in data center networks. This paper compares the Dramp algorithm with the equal-cost multipath (ECMP) and global first fit (GFF) routing algorithms on the Mininet simulation platform. The results show that Dramp gives the best performance; thus, Dramp is simpler and more practical for traffic engineering in data center networks.

Key words：network congestion software defined network (SDN)degree of critical link multipath routing traffic balance

收稿日期: 2015-06-23 出版日期: 2016-04-01

ZTFLH:

TP393.2

通讯作者:杨家海,研究员,E-mail:yang@cernet.edu.cnE-mail: yang@cernet.edu.cn

引用本文:

杨洋, 杨家海, 秦董洪. 数据中心网络多路径路由算法[J]. 清华大学学报（自然科学版）, 2016, 56(3): 262-268.
YANG Yang, YANG Jiahai, QIN Donghong. Multipath routing algorithm for data center networks. Journal of Tsinghua University(Science and Technology), 2016, 56(3): 262-268.

链接本文:

http://jst.tsinghuajournals.com/CN/10.16511/j.cnki.qhdxxb.2016.21.031或 http://jst.tsinghuajournals.com/CN/Y2016/V56/I3/262

图表:

图1　原型系统

图2　FatTree(K=8)

图3　实验拓扑

图4　时延比较

图5　丢包率比较

图6　吞吐量比较

参考文献:

[1] Arsalan T, Martin C, Teemu K, et al. Applying nox to the datacenter[C]//Proceedings of workshop on Hot Topics in Networks (HotNets-VIII). New York, NY, USA:ACM, 2009.
[2] Theophilus B, Ashok A, Aditya A, et al. MicroTE:Fine grained traffic engineering for data centers[C]//CoNEXT'11 Conference on emerging Networking Experiments and Technologies. New York, NY, USA:ACM, 2011, 8.
[3] Mohammad A, Sivasankar R, Barath R, et al. Hedera:Dynamic flow scheduling for data-center networks[C]//NSDI'10 Proceedings of the 7th USENIX conference on Networked systems design and implementation. Berkeley, CA, USA:USENIX Association, 2010:19-19.
[4] Andrew R, Wonho K, Praveen Y. Mahout:Low-overhead datacenter traffic management using end-host-based elephant detection[C]//INFOCOM'11 Proceedings IEEE INFOCOM 2011. Shanghai, China:IEEE, 2011:1629-1637.
[5] Jonathan P, Amy O, Hari B, et al. Fastpass:A centralized zero-queue datacenter network[C]//SIGCOMM'14 Proceedings of the 2014 ACM conference on SIGCOMM. New York, NY, USA:ACM, 2014:307-318.
[6] Andrew R, Jeffrey C, JeanT, et al. DevoFlow:Scaling flow management for high-performance networks[C]//SIGCOMM'11 Proceedings of the 2014 ACM conference on SIGCOMM. New York, NY, USA:ACM, 2011:254-265.
[7] Yukihiro N, Kazuki H, Lee C, et al. DomainFlow:Practical flow management method using multiple flow tables in commodity switches[C]//CoNEXT'13 Proceedings of the ninth ACM conference on Emerging networking experiments and technologies. New York, NY, USA:ACM, 2013:399-404.
[8] Eric J, Deng P, Liu J, et al. Asimulation and emulation study of sdn-based multipath routing for fat-tree data center networks[C]//WSC'14 Proceedings of the 2014 Winter Simulation Conference. Piscataway, NJ, USA:IEEE Press, 2014:3072-3083.
[9] Albert G, James R H, Navendu J, et al. VL2:A scalable and flexible data center network[J]. Communications of the ACM, 2011, 54(3):95-104.
[10] Zhou J, Malveeka T, Zhu M, et al. WCMP:Weighted cost multipathing for improved fairness in data centers[C]//EuroSys'14 Proceedings of the Ninth European Conference on Computer Systems. New York, NY, USA:ACM, 2014, 5.
[11] Pettit J. Open vSwitch. (2014-11-03). http://openvswitch.org/pipermail/announce/2014-December/000071.html.

No related articles found!