logo

SCIENTIA SINICA Informationis, Volume 48, Issue 3: 293-314(2018) https://doi.org/10.1360/N112017-00203

An overview of software-defined network measurement technologies

More info
  • ReceivedOct 25, 2017
  • AcceptedFeb 28, 2018
  • PublishedMar 16, 2018

Abstract

In recent years, with the continuous expansion of the size of networks, the complexity of network structure and functionality and the volume of network traffic have immensely increased. This has resulted in severe challenges in configuring, operating and managing traditional networks. As a prominent way of monitoring and handling network status in order to optimize network structure, enhance the quality of service and achieve network error detection and recovery, software-defined networks, as a new emerging network structure, has considered all of these problems and suits these network applications well. This paper is a survey of the state-of-the-art research on measurement technologies in software-defined networks, of which we detail the design concept, performance approach, measurement objects, technical advantages and disadvantages from three standpoints: network performance, topology, and traffic measurement. We mainly cover latency, packet loss and bandwidth measurements when considering network performance measurement. Finally, the technological approaches are summarized and future research issues are discussed.


Funded by

国家自然科学基金(61379145)

国家自然科学基金(61702539)


Acknowledgment

感谢审稿专家的中肯建议及编委的辛勤工作


References

[1] Kreutz D, Ramos F M V, Esteves Verissimo P, et al. Software-defined networking: a comprehensive survey. Proc IEEE, 2014, 103: 10--13. Google Scholar

[2] Cai Z P, Yin J P, Liu F, et al. Efficiently monitoring link bandwidth in IP networks. In: Proceedings of IEEE Global Telecommunications Conference, St. Louis, 2005. Google Scholar

[3] Yan X C, Yin J P, Cai Z P. Survey on network topology discovery algorithm. Comput Eng Appl, 2007, 43: 131--135. Google Scholar

[4] Tsai P W, Hsu C Y, Luo M Y. On the Implementation of Adaptive Flow Measurement in the SDN-enabled Network: A Prototype. APAN Proc, 2015, 40: 7 CrossRef Google Scholar

[5] Akyildiz I F, Lee A, Wang P, et al. A roadmap for traffic engineering in SDN-OpenFlow networks. Int J Comput Telecommun Netw, 2014, 71: 1--30. Google Scholar

[6] Zhou T Q, Cai Z P, Xia J, et al. Traffic engineering for software defined networks. J Softw, 2016, 27: 394--417. Google Scholar

[7] Nunes B A A, Mendonca M, Nguyen X N. A Survey of Software-Defined Networking: Past, Present, and Future of Programmable Networks. IEEE Commun Surv Tutorials, 2014, 16: 1617-1634 CrossRef Google Scholar

[8] Karakus M, Durresi A. Quality of Service (QoS) in Software Defined Networking (SDN): A survey. J Network Comput Appl, 2017, 80: 200-218 CrossRef Google Scholar

[9] Cai Z P. Network measurement technologies, models and algorithms based on active and passive measurement. Dissertation for Ph.D. Degree. Changsha: National University of Defense Technology, 2005. Google Scholar

[10] Yu C, Lumezanu C, Sharma A, et al. Software-defined latency monitoring in data center networks. In: Proceedings of Software-defined Latency Monitoring in Data Center Networks, New York, 2015. 360--372. Google Scholar

[11] Adrichem N L M V, Doerr C, Kuipers F A. OpenNetMon: network monitoring in OpenFlow software-defined networks. In: Proceedings of IEEE Network Operations and Management Symposium, Krakow, 2014. Google Scholar

[12] Sinha D, Haribabu K, Balasubramaniam S. Real-time monitoring of network latency in software defined networks. In: Proceedings of IEEE International Conference on Advanced Networks and Telecommuncations Systems, Kolkata, 2015. Google Scholar

[13] Atary A, Bremler-Barr A. Efficient round-trip time monitoring in OpenFlow networks. In: Proceedings of the 35th Annual IEEE International Conference on Computer Communications, San Francisco, 2016. Google Scholar

[14] Nguyen H N, Begin T, Busson A, et al. Evaluation of an end-to-end delay estimation in the case of multiple flows in SDN networks. In: Proceeding of the 12th International Conference on Network and Service Management, Montreal, 2017. 336--341. Google Scholar

[15] Panwaree P, Kim J, Aswakul C. Packet delay and loss performance of streaming video over emulated and real OpenFlow networks. In: Proceedings of the 29th International Technical Conference on Circuit/Systems Computers and Communications, At Phuket, 2014. Google Scholar

[16] Fu C, John W, Meirosu C. EPLE: an efficient passive lightweight estimator for SDN packet loss measurement. In: Proceedings of IEEE Conference on Network Function Virtualization and Software Defined Networks, Palo Alto, 2017. 192--198. Google Scholar

[17] John W, Meirosu C. Low-overhead packet loss and one-way delay measurements in service provider SDN. ONS, 2014. http://www.usenix.org/sites/default/files/ons2014-poster-john.pdf. Google Scholar

[18] Li Y, Miao R, Kim C, et al. LossRadar: fast detection of lost packets in data center networks. In: Proceedings of the 12th International on Conference on Emerging Networking Experiments and Technologies, Irvine, 2016. 481--495. Google Scholar

[19] Cheng P, Ren F Y, Shu R, et al. Catch the whole lot in an action: rapid precise packet loss notification in data centers. In: Proceedings of the 11th USENIX Conference on Networked Systems Design and Implementation, Seattle, 2014. 17--28. Google Scholar

[20] Chan Edmond W W, Chen A, Luo X P, et al. TRIO: measuring asymmetric capacity with three minimum round-trip times. In: Proceedings of the 7th Conference on Emerging Networking Experiments and Technologies, Tokyo, 2011. Google Scholar

[21] John W, Meirosu C. Low-overhead packet loss and one-way delay measurements in service provider SDN. ONS, 2014. http://www.sjalander.com/wolfgang/publications/ons2014-poster-john.pdf. Google Scholar

[22] Sezer S, Scott-Hayward S, Chouhan P K, et al. Are we ready for SDN? implementation challenges for software-defined networks. IEEE Commun Mag, 2013, 51: 36--43. Google Scholar

[23] Yassine A, Rahimi H, Shirmohammadi S. Software defined network traffic measurement: current trends and challenges. IEEE Instrum Meas Mag, 2015, 18: 42--50. Google Scholar

[24] Tootoonchian A, Ghobadi M, Ganjali Y. OpenTM: traffic matrix estimator for OpenFlow networks. In: Proceedings of the 11th International Conference on Passive and Active Measurement, Zurich, 2010. 201--210. Google Scholar

[25] Chowdhury S R, Bari M F, Ahmed R, et al. Payless: a low cost network monitoring framework for software defined networks. In: Proceedings of IEEE Network Operations and Management Symposium, Ahmed, 2014. Google Scholar

[26] Rasley J, Stephens B, Dixon C, et al. Planck: millisecond-scale monitoring and control for commodity networks. In: Proceedings of the 2014 ACM Conference on SIGCOMM, Chicago, 2014. Google Scholar

[27] Sekar V, Reiter M K, Willinger W, et al. cSamp: a system for network-wide flow monitoring. 2008. https://www.usenix.org/legacy/event/nsdi08/tech/full_papers/sekar/sekar_html/. Google Scholar

[28] Yu C, Lumezanu C, Zhang Y P, et al. Flowsense: monitoring network utilization with zero measurement cost. In: Proceedings of International Conference on Passive and Active Network Measurement, Hong Kong, 2013. Google Scholar

[29] Benson T, Anand A, Akella A, et al. MicroTE: fine grained traffic engineering for data centers. In: Proceedings of the 7th Conference on Emerging Networking Experiments and Technologies, Tokyo, 2011. Google Scholar

[30] Suh Junho, Kwon T T, Dixon C, et al. Opensample: a low-latency, sampling-based measurement platform for commodity SDN. In: Proceedings of the 34th International Conference on Distributed Computing Systems, Madrid, 2014. Google Scholar

[31] Yu M, Jose L, Miao R. Software defined traffic measurement with OpenSketch. In: Proceedings of the 10th USENIX Conference on Networked Systems Design and Implementation, Lombard, 2013. Google Scholar

[32] Charikar M, Chen K, Farach-Colton M. Finding frequent items in data streams. In: Proceedings of International Colloquium on Automata, Languages, and Programming, Turku, 2004. Google Scholar

[33] Cormode G, Muthukrishnan S. What's new: finding significant differences in network data streams. IEEE/ACM Trans Networking, 2005, 13: 1219-1232 CrossRef Google Scholar

[34] Li Y L, Miao R, Kim C, et al. FlowRadar: a better netflow for data centers. In: Proceedings of the 13th Usenix Conference on Networked Systems Design and Implementation, Santa, 2016. Google Scholar

[35] Liu Z X, Manousis A, Vorsanger G, et al. One sketch to rule them all: rethinking network flow monitoring with UnivMon. In: Proceedings of the 2016 ACM Conference on SIGCOMM, Florianopolis, 2016. Google Scholar

[36] Schweller R, Li Z, Chen Y. Reversible Sketches: Enabling Monitoring and Analysis Over High-Speed Data Streams. IEEE/ACM Trans Networking, 2007, 15: 1059-1072 CrossRef Google Scholar

[37] Huang Q, Jin X, Lee P P C, et al. SketchVisor: robust network measurement for software packet processing. In: Proceedings of the Conference of the ACM Special Interest Group on Data Communication, Los Angeles, 2017. Google Scholar

[38] Alipourfard O, Moshref M, Yu M L. Re-evaluating measurement algorithms in software. In: Proceedings of the 14th ACM Workshop on Hot Topics in Networks, Philadelphia, 2015. Google Scholar

[39] Moshref M, Yu M L, Govindan R, et al. DREAM: dynamic resource allocation for software-defined measurement. In: Proceedings of the 2014 ACM Conference on SIGCOMM, Chicago, 2014. Google Scholar

[40] Sun P, Yu M, Freedman M J. HONE: Joint Host-Network Traffic Management in Software-Defined Networks. J Netw Syst Manage, 2015, 23: 374-399 CrossRef Google Scholar

[41] Xu H, Yu Z, Qian C. Minimizing Flow Statistics Collection Cost Using Wildcard-Based Requests in SDNs. IEEE/ACM Trans Networking, 2017, 25: 3587-3601 CrossRef Google Scholar

[42] Bahnasy M, Idoudi K, Elbiaze H. OpenFlow and GMPLS Unified Control Planes: Testbed Implementation and Comparative Study. J Opt Commun Netw, 2015, 7: 301-308 CrossRef Google Scholar

[43] Yonghong F, Jun B, Jianping W. A dormant multi-controller model for software defined networking. China Commun, 2014, 11: 45-55 CrossRef Google Scholar

[44] Wang J L, Shou G C, Hu Y H, et al. A multi-domain SDN scalability architecture implementation based on the coordinate controller. In: Proceedings of International Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery, Chengdu, 2016. Google Scholar

[45] Liu C, Malboubi A, Chuah C N. OpenMeasure: adaptive flow measurement inference with online learning in SDN. In: Proceedings of IEEE Conference on Computer Communications Workshops, San Francisco, 2016. Google Scholar

[46] Farshad A, Georgopoulos P, Broadbent M, et al. Leveraging SDN to provide an in-network QoE measurement framework. In: Proceedings of IEEE Conference on Computer Communications Workshops, Hong Kong, 2015. Google Scholar

[47] Dacier M C, Dietrich S, Kargl F, et al. Network attack detection and defense — security challenges and opportunities of software-defined networking. Dagstuhl Rep, 2017, 6: 1--28. Google Scholar

[48] Saxena M, Rakesh K. A recent trends in software defined networking (SDN) security. In: Proceedings of the 3rd International Conference on Computing for Sustainable Global Development, New Delhi, 2016. Google Scholar

[49] Hong D K, Ma Y D, Banerjee S, et al. Incremental deployment of SDN in hybrid enterprise and ISP networks. In: Proceedings of the Symposium on SDN Research, Santa Clara, 2016. Google Scholar

[50] Canini M, Feldmann A, Levin D, et al. Panopticon: incremental deployment of software-defined networking. ACM Symposium on SDN Research, 2016. https://www.opennetworking.org/images/stories/downloads/sdn-resources/IEEE-papers/panopticon-incremental-deplyment-of-sdn.pdf. Google Scholar

  • Figure 1

    Research on SDN measurement classification

  • Figure 2

    Measurement phases on GRAMI technology

  • Figure 3

    Example of RTT measurement technology

  • Figure 4

    (Color online) Packet loss measurement of EPLE. MEP is the abbreviation of measurement end point, MIP is the abbreviation of measurement installation point

  • Figure 5

    (Color online) Structure design of LossRadar measurement

  • Figure 6

    (Color online) Structure of SketchVisor measurement

  • Figure 7

    (Color online) Framework of DREAM traffic measurement system

  • Figure 8

    Framework of HONE traffic measurement system

  • Table 1   Comparison of SDN network performance measurement technologies
    名称 测量对象 测量方式 技术特点 优缺点分析
    SLAM [10] 延迟 主动 充分利用OpenFlow协议功能,不需要对硬件进行额外设计与支持.测量准确性与开销成正相关. 测量思路简单新颖,在可能影响延迟测量结果准确性方面考虑不足,例如控制链路延迟的具体测法.
    TTL-based looping [12] 主动 利用TTL作为控制参数控制数据包的循环次数,从而获取链路的平均延迟. 思路清晰,实现难度较小.但测量结果的准确性和时效性不高,只能满足一般网络应用.
    GRAMI [13] 主动 预设测量点MP,计算覆盖网络(overlay network),实现网络中任意节点间延迟测量. 需要设计高效的MP选择算法,优化overlay计算效率,测量结果准确性高,但其网络故障恢复能力较弱,耗时较长.
    End-to-End Multiflow delay [14] 主动 计算路径上每段链路延迟并求和,获取端到端延迟. 通过严格的数学公式推导,得出期望与方差的数学关系,测量结果准确性高.对网络设备的计算性能要求很高,适用范围较小.
    EPLE [16,17] 丢包 被动 基于OpenFlow协议实现丢包检测.引入零探测流量,适用与轻量级的丢包测量. 测量开销为零,计算处理开销仅增加了4%$\sim$9%,内存开销仅增加了1.5%$\sim$3%.
    LossRadar [18] 主动 采用invertible Bloom filter编码方式减少内存资源占用,同时能根据需要灵活地提取丢包相关的流的信息. 测量方式简单,测量结果误差小.测量策略灵活,可按需获取出现丢包的流的相关信息.
    CP [19] 主动 改变传统丢包方式,只丢弃数据部分,报文头保留,以获得更好的测量结果. 测量方式新颖,缓冲区分块,以存放待转发数据包以及丢弃数据内容后的数据包头.对丢包信息的回传也可在一定程度上保护网络性能.
    ABWM [21,22] 带宽 被动 测量思想基于已有可用带宽测量的改进,适用于SDN网络架构. 测量前提是网络中每条链路带宽已知,且背景流波动在可控范围之内,测量条件较为理想,据实际应用还有差距.
    OpenNetMon [11]

    延迟

    丢包

    带宽

    主动 OpenNetMon的测量内容涵盖了延迟、丢包、带宽等主要网络性能指标,测量方案的设计思路具有创新性、灵活度高、扩展性好. 测量方法简单有效,测量结果的准确性和实时性相对较高.
  • Table 2   Comparison of SDN Traffic measurement technologies
    名称 测量对象 测量方式 技术特点 优缺点分析
    OpenTM [24]

    字节数

    持续时间

    主动 固定频率向交换机请求流的统计数据. 采样节点选取方式较主观,对测量结果有影响.结果准确性与开销成正相关.
    SketchVisor [37]

    流量速率

    数据包数

    链路利用率

    主动 提供快速通路实现高速流量处理性能. 克服Sketch设计本身缺陷,提高高速网络下流量处理的效率.测量结果的准确性相对下降,同时在流量分流的处理上,缺少灵活性,主动选择能力不足.
    PayLess [25] 主动 自适应频率对流信息进行采样. 根据应用具体需求和当前网络流量状态,动态调整对流的采样频率,减少了网络开销.
    Planck [26] 主动 利用商用交换机端口镜像功能,对流入流量进行备份并分析. 可对网络流进行全方位分析,结果可靠准确.但对端口处理性能要求高.需要考虑突发流情况下,端口的缓冲和应对机制.
    FlowSense [28] 被动 基于OpenFlow协议改进的流量测量技术. 测量结果准确,测量开销较低,适用与小流检测,在大流测量方面实时性不高.
    MicroTE [29] 被动 适用于小流检测,可提前感知流量突发情况,并提供重路由方案. 测量开销低,扩展性能强,需要对硬件升级,普及成本较高,不利于市场推广.
    OpenSample [30] 被动 适用于任意TCP流的检测,对Elephant流检测结果较理想. 改进sFlow采样技术,测量效率和延迟均得到有效降低,同时Elephant流检测具有较高市场价值.

Copyright 2019 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有

京ICP备18024590号-1