Re: [babel] Warren Kumari's Discuss on draft-ietf-babel-v4viav6-07: (with DISCUSS and COMMENT)

[Warren Kumari <[e-mail]>]

On Tue, Feb 15, 2022 at 10:39 AM Warren Kumari via Datatracker <
[e-mail]> wrote:

> Warren Kumari has entered the following ballot position for
> draft-ietf-babel-v4viav6-07: Discuss
>
> When responding, please keep the subject line intact and reply to all
> email addresses included in the To and CC lines. (Feel free to cut this
> introductory paragraph, however.)
>
>
> Please refer to https://www.ietf.org/blog/handling-iesg-ballot-positions/
> for more information about how to handle DISCUSS and COMMENT positions.
>
>
> The document, along with other ballot positions, can be found here:
> https://datatracker.ietf.org/doc/draft-ietf-babel-v4viav6/
>
>
>
> ----------------------------------------------------------------------
> DISCUSS:
> ----------------------------------------------------------------------
>
> I'd started balloting this as Abstain, but while writing up the ballot I
> realized that it's important enough that it deserves to be DISCUSSed.
>
> This is clearly clever, but feels to me like it might fall into "Oo, you
> are so
> sharp you’ll cut yourself one of these days"[0] territory.
>
> I'm not saying that the "v4-via-v6" is a *bad* idea, but I really don't
> think
> that it should be introduced / documented in a Standards Track Babel
> document -
> it touches core plumbing, and should be discussed and documented in a
> V6OPS (or
> 6MAN) document, and then this document includes it by reference.
>
> If this was only ever going to used in Babel environments I'd be much less
> concerned, but I suspect (hope?) that future solutions will want to do very
> similar things, and that it needs to be reviewed with an assumption that it
> might get widely used. It should documented in a "self contained" manner
> so it
> can be cleanly referenced - at the moment, a reference would need to point
> at
> bits of Section 1 and 3, and there is some feeling of "this is probably
> safe,
> the 192.0.0.8 bit might make operations / debugging a bit harder, but...
> ¯\_(ツ)_/¯"
>
> If this has already received significant discussion in V6OPS / similar, or
> if
> it is already clearly documented elsewhere[1], I'll clear my DISCUSS and
> Abstain or support it.
>
> I'm sure that this DISCUSS will be frustrating to the authors/WG - I'm
> doing so
> because I'd like to see this technique more able to be used (and make sure
> that
> there aren't any sharp pointy bits), not because I think it's a bad idea...
>
> [0]: quote from Terry Pratchett, Thief of Time
> [1]: I suspect it is already documented somewhere, but the closest I can
> think
> is RFC7600 - "IPv4 Residual Deployment via IPv6 - A Stateless Solution
> (4rd)",
> an Experimental document which is noticeably different to this. If it *is*
> already documented somewhere else though, then why is this not just
> referencing
> that instead?
>
>

Someone pointed out to me (off-list, probably in an attempt to spare me
embarrassment on the assumption that I'm not already used to looking
foolish) that e.g BGP advertising v6 next hops for v4 does some similar.


That's entirely true, but I don't remember (and I'm on my phone at the
moment and cannot easily check) much discussion on things like how to deal
with things like sourcing packet too big, etc.

If this is all already covered and discussed elsewhere, I'm happy for the
document to point at that, and we'll be good...

W


> ----------------------------------------------------------------------
> COMMENT:
> ----------------------------------------------------------------------
>
> Thanks for a well written document.
>
> Other than my discuss, I only have a question:
> Section 5 (Backwards Compatibility) says:
> "As a result, incompatible versions will ignore v4-via-v6 routes. "
>
> Is it *always* safe for a babel router to ignore a route? I really haven't
> thought about it enough (and the fact that it is DV based makes me think
> that
> it should be fine) but I'd like some reassurance that it is, especially in
> the
> case that a prefix is originated by multiple routers, and one of them gets
> filtered/ignored.
>
>
>
> --
Perhaps they really do strive for incomprehensibility in their specs.
After all, when the liturgy was in Latin, the laity knew their place.
-- Michael Padlipsky