Re: [v6ops] Discussion focus: draft-ietf-v6ops-ipv6rtr-reqs

[Fred Baker <[e-mail]>]

Thanks for this. It's cogent and to the point.

I do have a question. The authors are from LinkedIn and Comcast. How would you interpret that fact in the context of the facts you point out?

> On Feb 5, 2018, at 7:34 AM, STARK, BARBARA H <[e-mail]> wrote:
> 
>>> The primary focus is routers for enterprise or carrier use, however
>>> most of the features also have broad applicability to all routers.
> 
> I see a real need for recommendations for enterprise routers. I also believe that any attempts to broaden the scope beyond that will both slow down producing such recommendations and force the recommendations to be watered down to the point whether they are significantly less useful to enterprise networks.
> Here is why I think most other router types either don't need or will ignore these recommendations:
> - CE routers are covered by RFC 7084
> - telco ISPs aren't asking for this (for their own ISP networks) and tend to use BBF documents; any who are waiting until IETF produces router requirements for them need to reconsider their strategy
> - cable ISPs aren't asking for this (for their own ISP networks) and tend to use CableLabs docs; any who are waiting until IETF produces router requirements for them need to reconsider their strategy
> - wireless ISPs aren't asking for this and tend to use 3GPP docs; any who are waiting until IETF produces router requirements for them need to reconsider their strategy
> - transit/core network providers aren't asking for this and have already got IPv6 running (and if they haven't, there's no excuse, because the equipment is there); they also don't tend to do address assignment (no hosts on those networks) so requirements for DHCPv6 and SLAAC aren't needed; a lot of this equipment is done with SDN now; this is also true of a lot of the "regional network" part of ISP networks
> - IXCs don't tend to do address assignment and seem to have IPv6 in place; any that don't have IPv6 -- it's not for lack of requirements; a lot of them have gone to SDN
> 
> As for SDN: If you think an open source project doing SDN code is missing some IPv6 functionality, it might be easier to contribute the code, rather than write an RFC and see if some random person says "OMG, the IETF is saying do this; no-one else is writing this code so I guess I better do it".
> 
> We need to get enterprises transitioning. Whatever IETF can do to help that specific audience would be appreciated.
> Barbara
> 
> _______________________________________________
> v6ops mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/v6ops