Hi Les, */* Last email on this thread. It is clear that some people are so stuck here to their beliefs that no matter what others say it will not change their mind. */*
*> As you see, a single application (Flex) can support multiple “topologies” – but that is orthogonal.* The entire topic is to make support of multiple topologies easy - and it is not orthogonal. - One option is to signal its metrics with unique bits - one per topology. - The other option is to signal metrics to be applicable to any app which chooses to use it then trim the pool with affinities. The subject draft is about latter. I was (also) suggesting the former. Currently what seems to be the case is operationally not user friendly mix of both to build arbitrary topology. The end, Robert. On Mon, Aug 23, 2021 at 8:24 PM Les Ginsberg (ginsberg) <ginsb...@cisco.com> wrote: > Robert – > > > > > > *From:* Robert Raszuk <rob...@raszuk.net> > *Sent:* Monday, August 23, 2021 10:44 AM > *To:* Les Ginsberg (ginsberg) <ginsb...@cisco.com> > *Cc:* Shraddha Hegde <shrad...@juniper.net>; Tony Przygienda < > tonysi...@gmail.com>; Ron Bonica <rbon...@juniper.net>; lsr@ietf.org > *Subject:* Re: [Lsr] New Version Notification for > draft-hegde-lsr-asla-any-app-00.txt > > > > > > You continue to confuse Flex and ASLA. > > > > I don't think so ... This thread is about ASLA encoding for Flex-Algo. > > *[LES:] Noooo --- please read your own draft. **😊**]* > > *The work “Flex” does not appear anywhere in the draft – nor should it.* > > > > ASLA is an architecture that supports multiple applications – of which > Flex is just one of the supported applications. > > > > Care to elaborate why ASLA SABM allows only 64 applications ? Even MT > allowed 4K topologies. > > > > *[LES:] Application != topology* > > *As you see, a single application (Flex) can support multiple “topologies” > – but that is orthogonal.* > > *The real issue here is you continue to not fully understand the > distinction between an “application” and the capabilities of a given > application – one of which may be to support multiple “topologies”.* > > Flex is a specific application which supports multiple “sets of > constraints” – AKA algorithms. > > > > With that definition everything is Flex. Topologies differ one from > another by running SPT on a different sets of links with different "sets of > constraints" > > > > *[LES:] Nooo…the bits in the SABM mask are assigned in > https://www.iana.org/assignments/igp-parameters/igp-parameters.xhtml#link-attribute-application-identifiers > <https://www.iana.org/assignments/igp-parameters/igp-parameters.xhtml#link-attribute-application-identifiers>* > > *(Note that Flex has a bit assignment which is not yet reflected there.)* > > *You need to allow that a new application might have completely semantics. > * > > And, Flex is doing just fine as it is – one of the indicators of which I > believe is your enthusiasm for it. 😊 > > > > Let's see if the reverse works :) > > > > *[LES:] Sorry – standing the hierarchy on its head does not make sense – > no matter how many times you suggest it.* > > > > * Les* > > > > Best, > > R, >
_______________________________________________ Lsr mailing list Lsr@ietf.org https://www.ietf.org/mailman/listinfo/lsr
