[v6ops] Re: 答复: Re: 答复: Re: Call For Adoption: draft-link-v6ops-6mops

[Jen Linkova <[e-mail]>]

On Thu, Aug 8, 2024 at 9:47 PM Aijun Wang <[e-mail]> wrote:
> No.  Declaring the “IPv6-Only” network does not preclude that the network can support IPv4 communication.
>
> The IPv4 communication can still be transmitted via the tunnel technology, that is “IPv4 Communication as a IPv6 service”

What can be run on top of tunnels doesn't matter here. What matters
is: does the network provides hosts with IPv4 addresses, IPv4 default
gateway and IPv4 native connectivity?
Actually when we discussed the  terminology for RFC8925, my initial
idea was to use 'IPv4-as-a-service' term, but someone (Jordi?) pointed
out that that term is already claimed and means exactly that - IPv4
over some tunnels, not "providing IPv4 only if the host needs it"

> Comparing with evolution of mobile technologies, there is no operator declare their network is “4G-mostly network” when they want to put forward to the 4G phase, but need still support the 3G host/endpoint.

I know very little about about mobile networks, but I got an
impression that introducing new technology in the mobile world seems
to be slightly easier.
I see mobile operators shutting down 3G networks (and if one of my
phones is still 3G only - it's my problem).

Turning off IPv4 in networks like public WiFi or enterprise
environments is a completely different story - the landscape is much
more diverse.

> Introducing the “IPv6-Mostly Network” concept, in my POV, is worse than the “Dual Stack”,

Wait, I was under - apparently wrong - impression that you have some
objections to the terminology, and you would prefer a different name
for a network where IPv6-only and IPv4-enabled clients co-exist.
Or do you actually object to the concept of such network, not just the
term we use?

> and it will also mask some hinder problems that can’t be emerged at the dual-stack phase.

I'm all ears! ;) My favorite section of the document is 'Typical
issues', would love to hear more war stories!

> When the operators declare clearly they are toward to the “IPv6-Only Network”, they can certainly accelerate the conversion of IPv6-only application,

Can they really?

>and also the gradual removal of the outdated hosts. But, introduce the concept of “IPv6-Mostly Network”, can give the transition more time to take action-----Similar with the effect of “Dual Stack”.

I'm not sure I see why you can't do it with IPv6-mostly and why
IPv6-mostly slows things down. The only thing this deployment model
does is allows you to have a single network segment for both IPv6-only
and IPv4-enabled hosts, instead of separating them and ensuring that
IPv6-only capable hosts are using IPv6-only network, and hosts which
need IPv4 are using another one (an approach which has extremely high
operational costs, if you want to do it at scale).
Could you please elaborate on your comment (I think we tried to
explain benefits of IPv6-mostly in section 5, but the text might need
more work).

> 发件人: Tim Chown [mailto:[e-mail]]
> 发送时间: 2024年8月8日 18:55
> 收件人: Brian Carpenter <[e-mail]>; Aijun Wang <[e-mail]>
> 抄送: Ron Bonica <rbonica=[e-mail]>; IPv6 Ops WG <[e-mail]>
> 主题: Re: [v6ops] Re: 答复: Re: Call For Adoption: draft-link-v6ops-6mops
>
>
>
> Indeed, it’s a very nice way to get on the path to removing IPv4 and a nice term for that process; the “most” referring to the property that most hosts on a subnet switch to IPv6-only, while those not capable continue to use IPv4 for some or all operation.
>
>
>
> Tim
>
>
>
> From: Brian Carpenter <[e-mail]>
> Date: Thursday, 8 August 2024 at 11:51
> To: Aijun Wang <[e-mail]>
> Cc: Ron Bonica <rbonica=[e-mail]>, IPv6 Ops WG <[e-mail]>
> Subject: [v6ops] Re: 答复: Re: Call For Adoption: draft-link-v6ops-6mops
>
> Aijun,
>
>
>
> I disagree. When talking to site operators who want to proceed towards IPv6 infastructure but have vital systems that cannot be updated from IPv4 immediately, the new term "IPv6 mostly" is exactly what they want to hear. Many sites are in that situation but would like to avoid dual stack on the wire.
>
>
>
> (via tiny screen & keyboard)
> Regards,
>         Brian Carpenter
>
>
>
> On Thu, 8 Aug 2024, 22:33 Aijun Wang, <[e-mail]> wrote:
>
> Hi,
>
> It seems that the newly assigned name "IPv6-Mostly network" may lead confusion or need more explanations to the customers.
> How about change the document name solely to "Deployment and Operations Consideration on IPv6-Only Network", and omit the introduction of new concept of "IPv6-Mostly network"?
>
> And, for the operator transit to IPv6-Only network, besides the C2S(client to server) communication, the C2C(client to client) communication requirement should also need to be addressed. It seems the document is lack of consideration for such part.
>
>
> Best Regards
>
> Aijun Wang
> China Telecom
>
> -----邮件原件-----
> 发件人: [e-mail] [mailto:[e-mail]] 代表 Jen Linkova
> 发送时间: 2024年8月5日 22:22
> 收件人: Ron Bonica <rbonica=[e-mail]>
> 抄送: [e-mail]
> 主题: [v6ops] Re: Call For Adoption: draft-link-v6ops-6mops
>
> The draft can be found at
> https://datatracker.ietf.org/doc/draft-link-v6ops-6mops/
>
> On Mon, Aug 5, 2024 at 11:21 PM Ron Bonica <rbonica=[e-mail]> wrote:
> >
> > Friends,
> >
> > This message begins a Call For Adoption for draft-link-v6ops-6mops. Please read the draft and send your comments in response to this email.
> >
> > The call for adoption will close on August 19, 2024.
> >
> >                                                               Ron
> >
> >
> >
> > Juniper Business Use Only
> >
> > _______________________________________________
> > v6ops mailing list -- [e-mail]
> > To unsubscribe send an email to [e-mail]
>
>
>
> --
> Cheers, Jen Linkova
>
> _______________________________________________
> v6ops mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
>
> _______________________________________________
> v6ops mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]
>
> _______________________________________________
> v6ops mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]



-- 
Cheers, Jen Linkova