Increasing End-to-End Availability over Multiple Autonomous Systems


This paper presents a coordinated approach for increasing the end-to-end availability of inter-domain paths. Major reasons for intermittent failures are frequent inter-domain route changes and their long convergence times. The first part of our solution reduces the need for inter-domain path changes by introducing the Resilient Chain. The two other mechanisms — Fast Scoped Rerouting and Virtual Convergence — are intended to limit the impact of BGP rerouting. Keywords— Inter-domain, BGP, Availability, Routing, Convergence

5 Figures and Tables

Cite this paper

@inproceedings{Engel2005IncreasingEA, title={Increasing End-to-End Availability over Multiple Autonomous Systems}, author={Thomas Engel}, year={2005} }