Characterizing Inter-domain Rerouting after Japan Earthquake

The Internet is designed to survive when some of its components become faulty. Rerouting is an important approach to bypassing faulty links. In this paper, we propose a method to characterize inter-domain rerouting as a result of the massive earthquake in Japan on March 2011. Moreover, the characterization helps correlate inter-domain rerouting events and end-to-end path-quality degradation measured in Hong Kong. We analyse the variation of AS betweenness centrality extracted from BGP data to identify the time span when most routes changed, the ASes that were most seriously affected, and the correlative and backup paths. The results show that three major providers of inbound traffic to Hong Kong were affected by unstable routing state caused by a cable fault after the earthquake. Our work provides a new method of utilizing control plane's information to diagnose data plane's problem.

[1]  Shunji Abe,et al.  Impact of Tohoku earthquake on R&E network in Japan , 2011, SWID '11.

[2]  Zhen Wu,et al.  BGP routing dynamics revisited , 2007, CCRV.

[3]  Jun Li,et al.  I-seismograph: Observing and measuring Internet earthquakes , 2011, 2011 Proceedings IEEE INFOCOM.

[4]  R. Bush,et al.  The Japan earthquake: the impact on traffic and routing observed by a local ISP , 2011, SWID '11.

[5]  Leonard M. Freeman,et al.  A set of measures of centrality based upon betweenness , 1977 .

[6]  Marco Chiesa,et al.  Analysis of country-wide internet outages caused by censorship , 2011, IMC '11.

[7]  Jiawei Han,et al.  Data Mining: Concepts and Techniques , 2000 .

[8]  Petter Holme.  Edge overload breakdown in evolving networks. , 2002, Physical review. E, Statistical, nonlinear, and soft matter physics.

[9]  Kuai Xu,et al.  Inferring Major Events from BGP Update Streams Technical Report 04-043 , 2004 .

[10]  Lixin Gao.  On inferring autonomous system relationships in the internet , 2001, TNET.

[11]  L. da F. Costa,et al.  Characterization of complex networks: A survey of measurements , 2005, cond-mat/0505185.

[12]  Fabián E. Bustamante,et al.  Distributed systems and natural disasters: BitTorrent as a global witness , 2011, SWID '11.

[13]  Beom Jun Kim,et al.  Vertex overload breakdown in evolving networks. , 2002, Physical review. E, Statistical, nonlinear, and soft matter physics.

[14]  Joan Feigenbaum,et al.  Learning-based anomaly detection in BGP updates , 2005, MineNet '05.

[15]  Xiapu Luo,et al.  Design and Implementation of TCP Data Probes for Reliable and Metric-Rich Network Path Monitoring , 2009, USENIX Annual Technical Conference.