Toward Programmable Interdomain Routing

End-to-end route control spanning a set of autonomous systems (ASes) can provide opportunities to both end users to optimize interdomain control and network service providers to increase business offerings. BGP, the de facto interdomain routing protocol, and recent interdomain proposals provide limited mechanisms for such control. We provide the first, systematic formulation of the software-defined internetworking (SDI) model, where an AS exposes a programmable interface to allow clients to define the interdomain routes of the network, and maintains its autonomy, by keeping the control of its export policies, to avoid fundamental violations such as valley routing. We develop a blackbox optimization algorithm to quickly find optimal export-policy-compliant end-to-end routes in SDI, and validate its efficacy using real interdomain topology. To understand the operational implication of SDI, we evaluate the privacy leakage brought by exposing an AS' available interdomain routes. Preliminary results show that a small number of neighbors or a large number of exposed RIB samples allows accurate inference on an AS' BGP selection policy, indicating a potential risk of not only SDI, but the whole interdomain routing community.1

[1]  Lixin Gao,et al.  Stable Internet routing without global coordination , 2000, SIGMETRICS '00.

[2]  Thomas E. Anderson,et al.  One tunnel is (often) enough , 2014, SIGCOMM.

[3]  Nando de Freitas,et al.  A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning , 2010, ArXiv.

[4]  João L. Sobrinho,et al.  Network routing with path vector protocols: theory and applications , 2003, SIGCOMM '03.

[5]  Scott Shenker,et al.  Privacy-Preserving Interdomain Routing at Internet Scale , 2017, Proc. Priv. Enhancing Technol..

[6]  Jennifer Rexford,et al.  MIRO: multi-path interdomain routing , 2006, SIGCOMM 2006.

[7]  Marco Canini,et al.  An Industrial-Scale Software Defined Internet Exchange Point , 2016, USENIX Annual Technical Conference.

[8]  Nando de Freitas,et al.  Taking the Human Out of the Loop: A Review of Bayesian Optimization , 2016, Proceedings of the IEEE.

[9]  Nick Feamster,et al.  SDX-Based Flexibility or Internet Correctness?: Pick Two! , 2017, SOSR.

[10]  Barry O'Sullivan,et al.  Constraint acquisition , 2017, Artif. Intell..

[11]  Gordon T. Wilfong,et al.  The stable paths problem and interdomain routing , 2002, TNET.

[12]  Robert Raszuk,et al.  Dissemination of Flow Specification Rules , 2009, RFC.

[13]  Russell J. Clark,et al.  SDX , 2014 .

[14]  Charu C. Aggarwal,et al.  Neural Networks and Deep Learning , 2018, Springer International Publishing.

[15]  Bernhard Ager,et al.  Outsourcing the routing control logic: better internet routing based on SDN principles , 2012, HotNets-XI.

[16]  Gregory N. Hullender,et al.  Learning to rank using gradient descent , 2005, ICML.

[17]  Yakov Rekhter,et al.  A Border Gateway Protocol 4 (BGP-4) , 1994, RFC.

[18]  Scott Shenker,et al.  Routing as a Service , 2006 .