Re: [v6ops] Making RDNSS a MUST?

[Tim Chown <[e-mail]>]

> On 3 Apr 2017, at 18:21, Ted Lemon <[e-mail]> wrote:
> 
> My reaction to this long and very predictable high-bandwidth back-and-forth is that I suspect the outcome will be the same as it was last time: no action.
> 
> This is frustrating.   It seems clear to me that requiring support for RDNSS would improve the situation.   Possibly requiring support for DHCPv6 would also improve the situation if it were done correctly (which is not a given).   But whenever anybody proposes requiring RDNSS, we get this demand for parity, and then a giant explosion of opinions, and then no action.
> 
> Essentially what I see here is a viable solution being held hostage to demands for a controversial solution.   It may be that the controversy isn't valid, and that really requiring DHCPv6 is the right thing to do.   But the controversy exists, and makes requiring DHCPv6 problematic.
> 
> I don't see any justification for a technical objection to RDNSS.   So my personal opinion is that despite the loud protests to the contrary, there is consensus to require RDNSS.   It may also be the case that there is consensus to require DHCPv6, on the same basis.   However, that would require a document that describes what "require DHCPv6" actually means.
> 
> It is of course entirely possible that the document authors are just plotzing at this discussion, or perhaps have chosen to ignore it, since the document isn't explicitly mentioned anywhere in the thread.   It would be valid for them to say that this is simply not an opportunity to re-open this issue.   However, the section in the document on Zero Touch Configuration certainly looks like a good place to insert a requirement.
> 
> My personal preference would be that if there is any serious desire for DHCPv6 to be required, that the proponents of this requirement write up in a document what that support looks like.   I don't think it makes sense to do that in this document.   I think it does make sense to require RDNSS in this document, because I don't think there's any question as to what that looks like.
> 
> I realize that holding that hostage probably looks like a great way for proponents of DHCPv6 support to get their way, but I don't think that's what would happen: what would happen would be that the authors of the document would say nothing at all on this topic.   So that's your choice: improve the situation, but don't get everything you want, or get nothing.   For those who do not want RDNSS, of course, the scorched earth approach will be fine, but for the rest of us I think it's not fine.


Ted, I don’t disagree; while there are some very good points made on both "sides”, a “no action” outcome doesn’t move us forward.

A reminder that “make support for RDNSS a MUST” is an open issue in RFC6434-bis (draft-clw-rfc6434-bis-01), as well as the Requirements for IPv6 Routers draft (draft-ali-ipv6rtr-reqs-02). 

Tim