(WG-chair hat firmly off) I am working on getting an analysis of the various fast-reroute technologies (MRT, not-via, optimal, Remote LFA) on different networks. Specifically, I want to update draft-enyedi-rtgwg-mrt-frr-algorithm with the method and results of analysis so that we can draw some conclusions about which MRT algorithm to select for unicast fast-reroute and how much difference the selection of the GADAG root makes.
To do this, I could use as many real network topologies as possible. While it's interesting to analyze arbitrary topologies, what is really meaningful is looking at the topologies that are used. For a network topology, I need a list of routers (router-name, IP-addresses) and a list of one-way links (local-router-name, local-interface-name, remote-route-name, remote-interface-name, IGP metric). The IP-addresses and router-names just have to be unique; what they are doesn't matter (IP addresses slightly for tie-breaking, but very minorly). I can also automatically translate the output of 'show isis database detail' and such, via Cariden. If you have such a model and would not be adverse to privately sharing it, with the understanding that no identifying information would be included or shared in the summarized analysis, I'd greatly appreciate it. Obviously, I have some networks that I'm looking at already, but the more the better the results. I'm also happy to discuss the methods that I'm using the compare the algorithms and their results. Thanks, Alia
_______________________________________________ rtgwg mailing list [email protected] https://www.ietf.org/mailman/listinfo/rtgwg
