Abstract
BGP allows routers to use general preference policies for route selection. This paper studies the impact of these policies on convergence time. We first describe a real-time model of BGP. We then state and prove a general theorem providing an upper bound on convergence time. Fi- nally, we show how to the use the theorem to prove conver- gence and estimate convergence time in three case studies.

This publication has 7 references indexed in Scilit: