Hi Jeff, On Mon, Jan 5, 2015 at 3:12 PM, Jeffrey Haas <[email protected]> wrote: > Robin, > > On Tue, Dec 23, 2014 at 07:18:14PM +0000, Lizhenbin wrote: >> Now multicast is an important service in the network such as IPTV, etc. >> But there are complex choice for multicast solutions including L2 >> multicast vs. L3 multicast, public multicast vs. multicast VPN, multicast >> based on PIM/BGP/mLDP/P2MP TE/IGP/Ingress replication which make the >> multicast provision difficult. This may be simplified through the I2RS >> intefaces. Could we define the multicast usecase in I2RS WG? And how will >> we define the multicast usecase: enhancing the >> draft-ietf-i2rs-usecase-reqs-summary-00 to incorporate multicast use cases >> or proposing independent multicast use case draft? > > Multicast is currently not in charter and our AD has given strong guidance > that we need to finish existing work before we pick up anything new. That > said, nothing stops you or others from creating an individual submission to > discuss this work. It may be some time before such a submission could be > adopted by the group though. > > Speaking as an individual contributor, multicast provides a significant > number of interesting problems that all must be solved in order to provide > an I2RS-like interface. For example, these might include: > - Interaction with interface/port yang modules. > - Definition of OIF-lists in YANG. > - Prefix filters (already a coordination issue IETF-wide) for managing > traffic filtering. > - Interaction with underlying layer 2 technologies for that layer multicast. > (And the requisite yang that would have to describe layer 2.) > - Interactions with MPLS technologies, most of which are still very new in > YANG or work has yet to even begin. > > While I believe such work is tractable, it is *quite* a bit of work. It > does have the interesting property that such work may drive requirements for > several other technologies. > > If you feel up to that particular challege this early in IETF Yang work, I > look forward to seeing your drafts.
I have no issue on this decision. You mentioned YANG three times above, that's where I have a question. I checked i2rs charter, it does not even mention YANG. My question is why YANG models became so essential, is it because they will be the way i2rs solution is developed? How could this be a solution by rewriting the protocol, e.g. bgp procedures in YANG format? A followup question: how can IESG evaluate a thousand or more line YANG program? I'll appreciate if you could shed some light into this. Regards, Behcet > > -- Jeff > > _______________________________________________ > i2rs mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/i2rs _______________________________________________ i2rs mailing list [email protected] https://www.ietf.org/mailman/listinfo/i2rs
