Rethinking iBGP Routing

Iuniana Oprescu, Mickael Meulle, Steve Uhlig, Cristel Pelsser, Olaf Maennel, Philippe Owzarski

The Internet is organized as a collection of administra- tive domains, known as Autonomous Systems (ASes). These ASes interact through the Border Gateway Protocol (BGP) that allows them to share reachability information. Adjacent routers in distinct ASes use external BGP (eBGP), whereas in a given AS routes are propagated over internal BGP (iBGP) sessions between any pair of routers. In large ASes where a logical full-mesh is not possible, confederations or route re- flectors (RRs) are used. However, these somewhat scalable alternatives have introduced their own set of unpredictable effects (persistent routing oscillations and forwarding loops causing an increase of the convergence time) extensively ad- dressed in the literature. The solution we propose to these issues consists of a struc- tured routing overlay holding a comprehensive view of the routes. We describe the design of a distributed entity that performs BGP route pre-computation for its clients inside a large backbone network and propagates the paths to the routers. Compared to the current iBGP routing, the advan- tage of the overlay approach is the separation between the responsibility of the control plane (route storage and best path computation) and the forwarding of the packets. One of the major improvements we bring is the divided routing table tackling the scalability concerns and allowing for parallel computation of paths.

In SIGCOMM 2010 Poster/Demo, August 2010.

Back to Cristel Pelsser's homepage