流追踪：一种软件定义网络中低开销的时延测量和路径追踪方法

Abstract

为了针对不同的应用和流量提供服务质量保障和差异化服务，负载均衡和多优先级队列技术被广泛地应用于网络中。在传统网络中，网络管理员经常使用“traceroute”和“ping”工具来检测负载均衡机制或者服务质量策略是否正常工作。然而，由于这些工具并不被现有的OpenFlow交换机所支持，所以还不能够应用于软件定义网络中。此外，traceroute和ping依靠主动发送探测包来探测路径。然而，当负载均衡机制把探测包和所需追踪流的数据包均衡到不同路径时，这些工具将无法探测出流的真实转发路径，更无法测量出真实的路径时延。因此，为了准确的测量链路时延，测量工具必须能够提前找出数据包的真实转发路径。基于此发现，我们提出了一套新的软件定义网络中的流追踪机制“FlowTrace”，利用它来追踪任意流量的转发路径以及测量数据流所经历的链路时延。该工具通过收集交换机的流表来计算流的转发路径。然而，如果直接从交换机中查询流表会产生很大的数据平面流量，从而带来巨大的开销。因此，我们提出了一种被动的零开销的流表收集方法来解决该问题。在获得流的真实转发路径后，我们提出了一种新的测量方法来测量不同流的网络时延。最后，实验结果显示我们设计的方法可以准确的找出流的真实转发路径并测量出不同种类流所经历的时延。

Keywords

软件定义网络 / 网络检测 / 路径追踪

References

[1] Agarwal, K., Rozner, E., Dixon, C., , 2014. SDN traceroute: tracing SDN forwarding without changing network behavior. Proc. 3rd Workshop on Hot Topics in Software Defined Networking, p.145–150.

[2] Al-Fares, M., Loukissas, A., Vahdat, A., 2008. A scalable, commodity data center network architecture. ACM SIGCOMM Comput. Commun. Rev., 38(4):63–74.

[3] Al-Fares, M., Radhakrishnan, S., Raghavan, B., , 2010. Hedera: dynamic flow scheduling for data center networks. Proc. 7th USENIX Conf. on Networked Systems Design and Implementation, p.19.

[4] Alizadeh, M., Yang, S., Sharif, M., , 2013. pFabric: minimal near-optimal datacenter transport. ACM SIGCOMM Comput. Commun. Rev., 43(4):435–446.

[5] Bai, W., Chen, L., Chen, K., , 2015. Informationagnostic flow scheduling for commodity data centers. Proc. 12th USENIX Conf. on Networked Systems Design and Implementation, p.455–468.

[6] Chowdhury, S.R., Bari, M.F., Ahmed, R., , 2014. Pay-Less: a low cost network monitoring framework for software defined networks. Proc. Network Operations and Management Symp., p.1–9.

[7] Clos, C., 1953. A study of non-blocking switching networks. Bell Syst. Tech. J., 32(2):406–424.

[8] Curtis, A.R., Kim, W., Yalagandula, P., 2011. Mahout: low-overhead datacenter traffic management using endhost-based elephant detection. Proc. IEEE INFOCOM, p.1629–1637.

[9] Ding, J., Huang, T., Liu, J., , 2015. Virtual network embedding based on real-time topological attributes. Front. Inform. Technol. Electron. Eng., 16(2):109–118.

[10] Greenberg, A., Hamilton, J.R., Jain, N., , 2009. VL2: a scalable and flexible data center network. ACM SIGCOMM Comput. Commun. Rev., 39(4):51–62.

[11] Guo, C., Yuan, L., Xiang, D., , 2015. Pingmesh: a large-scale system for data center network latency measurement and analysis. ACM SIGCOMM Comput. Commun. Rev., 45(4):139–152.

[12] Handigol, N., Heller, B., Jeyakumar, V., , 2012. Where is the debugger for my software-defined network? Proc. 1st Workshop on Hot Topics in Software Defined Networks, p.55–60.

[13] Jarschel, M., Zinner, T., Hohn, T., , 2013. On the accuracy of leveraging SDN for passive network measurements. Proc. Telecommunication Networks and Applications Conf., p.41–46.

[14] Katta, N.P., Rexford, J., Walker, D., 2013. Incremental consistent updates. Proc. 2nd ACM SIGCOMM Workshop on Hot Topics in Software Defined Networking, p.49–54.

[15] Kazemian, P., Chang, M., Zeng, H., , 2013. Real time network policy checking using header space analysis. Proc. 10th USENIX Conf. on Networked Systems Design and Implementation, p.99–112.

[16] Khurshid, A., Zhou, W., Caesar, M., , 2012. VeriFlow: verifying network-wide invariants in real time. Proc. 1st Workshop on Hot Topics in Software Defined Networks, p.49–54.

[17] Liu, J., Huang, T., Chen, J., , 2011. A new algorithm based on the proximity principle for the virtual network embedding problem. J. Zhejiang Univ.-Sci. C (Comput. & Electron.), 12(11):910–918.

[18] McKeown, N., Anderson, T., Balakrishnan, H., , 2008. OpenFlow: enabling innovation in campus networks. ACM SIGCOMM Comput. Commun. Rev., 38(2):69–74.

[19] Perešíni, P., Kuzniar, M., Vasić N., , 2013. OF.CPP: consistent packet processing for OpenFlow. Proc. 2nd ACM SIGCOMM Workshop on Hot Topics in Software Defined Networking, p.97–102.

[20] Phemius, K., Bouet, M., 2013. Monitoring latency with OpenFlow. Proc. 9th Int. Conf. on Network and Service Management, p.122–125.

[21] Phemius, K., Thales, B.M., 2013. OpenFlow: why latency does matter. Proc. IFIP/IEEE Int. Symp. on Integrated Network Management, p.680–683.

[22] Qi, H., Shiraz, M., Liu, J., , 2014. Data center network architecture in cloud computing: review, taxonomy, and open research issues. J. Zhejiang Univ.-Sci. C (Comput. & Electron.), 15(9):776–793.

[23] Reitblatt, M., Foster, N., Rexford, J., , 2012. Abstractions for network update. ACM SIGCOMM Comput. Commun. Rev., 42(4):323–334.

[24] Scott, C., Wundsam, A., Raghavan, B., , 2014. Troubleshooting blackbox SDN control software with minimal causal sequences. Proc. ACM SIGCOMM Conf., p.1–12.

[25] Su, Z., Wang, T., Xia, Y., , 2014. FlowCover: low-cost flow monitoring scheme in software defined networks. Proc. IEEE Global Communications Conf., p.1956–1961.

[26] Tavakoli, A., Casado, M., Koponen, T., , 2009. Applying NOX to the datacenter. Proc. 8th ACM Workshop on Hot Topics in Networks, p.1-6.

[27] Wundsam, A., Levin, D., Seetharaman, S., , 2011. OFRewind: enabling record and replay troubleshooting for networks. Proc. USENIX Annual Technical Conf., p.29.

[28] Yu, C., Lumezanu, C., Zhang, Y., , 2013. FlowSense: monitoring network utilization with zero measurement cost. Proc. 14th Int. Conf. on Passive and Active Measurement, p.31–41.

[29] Yu, C., Lumezanu, C., Sharma, A., , 2015. Softwaredefined latency monitoring in data center networks. Proc. 16th Int. Conf. on Passive and Active Measurement, p.360–372.

[30] Yu, M., Jose, L., Miao, R., 2013. Software defined traffic measurement with OpenSketch. Proc. 10th USENIX Conf. on Networked Systems Design and Implementation, p.29–42.

[31] Zhang, H., Lumezanu, C., Rhee, J., , 2014. Enabling layer 2 pathlet tracing through context encoding in software-defined networking. Proc. 3rd Workshop on Hot Topics in Software Defined Networking, p.169–174.

RIGHTS & PERMISSIONS

Zhejiang University and Springer-Verlag Berlin Heidelberg