> Comments inline. I think my original comment describes the state of play: > "Just because you don't like/understand it doesn't necessarily mean it's > wrong.".
John -- Irrespectively, as you like to say -- again, calling me stupid isn't helping anything here. > > > The approach of A/B/C independently generating route to X only based > > > on local MAC learning works even if physical link failure detection > > > is not an option. And in most practical cases, X will have enough > > > active flows that it will cause local learning on A/B/C to hopefully > > > happen well before a failure of connection to A happens. In the > > > event that this is not the case, current scheme takes care of it via ESI > route. > > > > Then -- again, aliasing is covering a corner case, and I'm not certain > > of the value of this mechanism in the main draft. > > [JD] Unsupported assertion No, what we've had in this thread is -- "This is not a corner case, as it happens all the time with LAGS," followed by, "Any given host will have enough flows to prevent the long convergence time failure mode from happening," above. Which is it? > > Agreed -- Which is why this needs some brainstorming, rather than just > > adding something into BGP that is fragile in the first place. > > [JD] Unsupported assertion Are you actually reading the thread here, or just reacting emotionally? > [JD] This is a local node behavior option that is entirely consistent w/ > aliasing > as it is currently defined If a local node can solve the problem in a way that's cleaner than what aliasing provides, then why does the draft need aliasing in the first place? Why are we pushing mechanisms into BGP for a problem that can be solved locally more efficiently and more reliably? > > As you say -- > > > > > I would think that the closer a PE is to the source of information > > > that causes it to generate a route - like local MAC learning for MAC > > > route, or LAG link present for ESI route - the more accurate its > > > information will be. The farther the PE is from the source of > > > information - like a PE depending on another PE's route to generate > > > its own MAC route - convergence will be slower and complexity will > > > be > > higher to keep track of dependent state. > > > > As aliasing suffers from the same criticism, > > [JD] Unsupported assertion Your assertion that my assertion is unsupported is, itself, unsupported. John -- clearly you don't want to have a discussion, so there's no point in having a "nondiscussion" with you. I'll find Ravi at the next IETF, and we can have an intelligent conversation -- which is apparently only possible when you're not in the room. Russ _______________________________________________ BESS mailing list [email protected] https://www.ietf.org/mailman/listinfo/bess
