Hello Carl, Interesting comment.
Would you agree with me that in principle a topology is a topology is a topology and, therefore, that we should strive for a model that can be employed recursively at any level in such a (implied) hierarchy ? When I look at topology, i.e fundamentally at a graph, I wish to manipulate or exploit aspects of the properties represented by it. In principle it doesn't matter to me what it represents. I have the same primitives at my disposal to manipulate it regardless of what it represents . Obviously in practice I might want to check some identifier in/of the topology to ensure that it's not a national backbone I'm messing with when I intended to change an access link in my metro ;-) Do you think that's a reasonable perspective from which to continue this discussion ? pd On Apr 6, 2016, at 5:41 PM, "Carl Moberg (camoberg)" <[email protected]> wrote: > > There is a case that can be made for topology models capable of exposing > topologies both from the view of a participant node as well as an all-seeing > orchestrator/controller. > > -- > Carl Moberg > Technology Director, CVG > [email protected] > >> On Apr 6, 2016, at 11:33 PM, Sterne, Jason (Nokia - CA) >> <[email protected]> wrote: >> >> Exposing a model of a *network* seems like something more appropriate on a >> controller that sits above individual NEs and aggregates network wide >> information no ? >> Jason >> >> -----Original Message----- >> From: netmod [mailto:[email protected]] On Behalf Of EXT Doolan, Paul >> (Coriant - US/Irving) >> Sent: Wednesday, April 06, 2016 17:33 >> To: Lou Berger >> Cc: [email protected] >> Subject: Re: [netmod] YANG model classification? >> >> Not sure either captures the case where, in the same network, there are >> instances of the model on NEs and on the management systems. >> >> Does "both" cover that case ? >> >> pd >> >> There has already been discussion of the concept of a switch matrix (inside >> an NE) which can On Apr 6, 2016, at 3:31 PM, Lou Berger <[email protected]> >> wrote: >> >>> My personal view is either. .. >>> >>> >>> On April 6, 2016 4:15:08 PM "Carl Moberg (camoberg)" <[email protected]> >>> wrote: >>> >>>> >>>> Is the YANG model in draft-ietf-teas-yang-te-topo expected to be >>>> implemented in a network element, on a management system or perhaps either? >>>> >>>> -- >>>> Carl Moberg >>>> Technology Director, CVG >>>> [email protected] >>>> >>>>> On Apr 6, 2016, at 8:09 PM, Scharf, Michael (Nokia - DE) >>>>> <[email protected]> wrote: >>>>> >>>>> draft-ietf-netmod-yang-model-classification suggests that a YANG model is >>>>> either a "Network Element YANG Model" or a "Network Service YANG Model". >>>>> >>>>> How would draft-ietf-teas-yang-te-topo be classified according to that? >>>>> Thoughts? >>>>> >>>>> Thanks >>>>> >>>>> Michael >>>>> >>>>> _______________________________________________ >>>>> netmod mailing list >>>>> [email protected] >>>>> https://www.ietf.org/mailman/listinfo/netmod >>>> >>>> _______________________________________________ >>>> netmod mailing list >>>> [email protected] >>>> https://www.ietf.org/mailman/listinfo/netmod >>>> >>> >>> >>> _______________________________________________ >>> netmod mailing list >>> [email protected] >>> https://www.ietf.org/mailman/listinfo/netmod >> >> _______________________________________________ >> netmod mailing list >> [email protected] >> https://www.ietf.org/mailman/listinfo/netmod >> >> _______________________________________________ >> netmod mailing list >> [email protected] >> https://www.ietf.org/mailman/listinfo/netmod > _______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
