Putting BGP on the Right Path: A Case for Next-Hop Routing

February 16, 2010
Time: 2:00pm-3:00pm
EE Conference Room
Hosted by: Prof. Gil Zussman
Speaker: Michael Schapira, Yale University and UC Berkeley

Abstract

BGP is plagued by many serious problems (protocol divergence, software bugs, misconfigurations, attacks, and more). Rather than continuing to add mechanisms to an already complex protocol, or redesigning interdomain routing from scratch, we propose making BGP *simpler*. We advocate a transition from today's path-based routing to a solution where ASes select and export routes based only on the neighboring ASes. Next-hop routing leads to simpler router implementation and configuration, and naturally expresses bilateral business relationships. Next-hop routing provably achieves faster convergence and incentive compatibility, side-stepping two major problems with today's BGP. Next-hop routing also allows operators to achieve their traffic-engineering and security goals, and enables multipath routing for additional performance and reliability benefits.

Joint work with Yaping Zhu and Jennifer Rexford.


500 W. 120th St., Mudd 1310, New York, NY 10027    212-854-3105               
©2014 Columbia University