Re: [v6ops] Thoughts about wider operational input

[Nick Buraglio <[e-mail]>]

Multihoming is just one piece of a much larger impediment apparatus.
Barrier of entry is the umbrella, and it is exacerbated by lack of support
across the board on common, modern - and in some cases brand new platforms.
Even if we're talking about dual stack, which as much as I hate to admit,
is probably the only viable path forward for most environments, the support
for IPv6 is haphazard and inconsistent in very common applications and
operating environments (take docker, for example).

I think a question we should be asking is not only "How can we make
operators /want/ to deploy IPv6?" but also "How can IPv6 be elevated to at
least a peer in software and product development". Perhaps this is just an
artifact of "squatters rights", but in my day-to-day, I almost never see
software of product development embrace IPv6 as the first pass protocol.
This may not sound that important, but what it implies and perpetuates is
two fold:

It promotes the idea that IPv4 is the "default", and that IPv6 is an
afterthought.
It very clearly states that IPv6 is not a requirement.

Operating systems managed to get the prioritization in place. How can that
"IPv6 by default" or even better - protocol agnostic mentality - take the
next step up the trail into product and software development?  Again, I
really don't see this as a technical problem, per se, but perhaps some
official best practices for activities such as:

"Software development protocol best practices"
"Product development protocol best practices"



nb


On Tue, Mar 22, 2022 at 10:21 AM Gert Doering <[e-mail]> wrote:

> Hi,
>
> On Tue, Mar 22, 2022 at 11:41:19AM +0100, Toerless Eckert wrote:
> > Why would it not be a desirable feature for me as the
> > end-user to have an IPv6/IPv6 NAT on the router towards the 3 Internet
> connections to
> > which i multi-home ? Sure, we can probably list some downsides, but they
> need to be
> > weighted against the benefits for me as an end user. Not ?
>
> This being an IETF list, it is not allowed to find anything-NAT beneficial.
>
> OTOH, as an operator, I find this a viable option.
>
> The reason why I prefer "host has multiple GUAs" better, in theory, is
> that it gives the *user* more control on what he wants - like, "I can
> run my bittorrent via the cable ISP, and SSH out via LTE", by selecting
> the corresponding source IP = selecting the outgoing ISP.
>
> Now, in practice, source address selection for "there are multiple GUAs
> with different pros/cons attached" is lacking quite a bit.  And HNCP
> routers with source-based forwarding can not be bought.  So all the
> nice things do not work.
>
>
> Which means, "a router with GUAs inside, a DSL port and a LTE modem and
> NPT66 on both outside interfaces" is a really nice redundancy solution
> for basically all "mostly unmanaged" SME networks.
>
> You'd pass control on "what should go where?" from the host to the router,
> but given that hosts are not there yet, that will get the job done.
>
> Gert Doering
>         -- NetMaster
> --
> have you enabled IPv6 on something today...?
>
> SpaceNet AG                      Vorstand: Sebastian v. Bomhard, Michael
> Emmer
> Joseph-Dollinger-Bogen 14        Aufsichtsratsvors.: A. Grundner-Culemann
> D-80807 Muenchen                 HRB: 136055 (AG Muenchen)
> Tel: [phone]         USt-IdNr.: DE813185279
>
> _______________________________________________
> v6ops mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/v6ops
>