Re: 202505061814.AYC Re: 202505060938.AYC Re: IPv4 Leases

[Tom Beecher via NANOG <nanog () lists nanog org>]

Abraham-

EZIP is a proposal. It could not be used operationally today for a
multitude of reasons that those of us who actually do this stuff have tried
to point out to you, but you continue to hand wave and ignore.

There are lots of people who read this list and may be confused by your
writings and think EZIP is a functional solution they could use today. This
is, frankly, kinda of a dick move. EZIP is nothing more than the idea you
have written down in your draft, and that in 9 years since you first
proposed it, nobody has really cared about. (
https://mailarchive.ietf.org/arch/search/?q=%22draft-chen-ati-adaptive-ipv4-address-space%22
)



On Tue, May 6, 2025 at 11:14 PM Abraham Y. Chen <aychen () avinta com> wrote:

> Hi, Tom:
>
> 0)    Thank you for your input.
>
> 1)    You are very accurate in reciting the facts and events. However, I
> am afraid that your conclusions might be flawed, because you probably
> overlooked a couple subtle aspects behind the EzIP proposal:
>
>     A.    The primary purpose of the EzIP is to expand the size capacity
> of the CG-NAT, nothing more. Besides activating the long-reserved 240/4 to
> be used like 100.64/10 which has been proven not too challenging, there is
> no other technical hurdle. As you may recall, CG-NAT was deployed over
> quarter century ago despite the objections from IETF, anyway.
>
>     B.    EzIP will form an "Overlay Network" that can practically be
> detached from the current Internet, while the two can work together as long
> as the interface between them conforms with the established protocols.
> Then, what happens within the EzIP can not be influenced from the outside.
> This is nothing new, either. There were precedents in the past, such as the
> Internet was deployed over PSTN with the dial-up modem serving as one of
> the umbilical cords between the two. The IPv6 was built on top of of the
> IPv4. Of course, the former was a big success story, while the latter is
> still a WIP.
>
>     C.    There are much more than a handful devices that can support the
> EzIP operation. Please have a look at Reference group E. on page 17 of the
> below whitepaper:
>
>     https://avinta.com/gallery/DeterministicInternet-SPKR.pdf
>
>     D.    In addition, IPv4 Unicast Extension Project has been working on
> networking devices that support 240/4:
>
>
> https://datatracker.ietf.org/doc/html/draft-schoen-intarea-unicast-240-08
>
>     Note that the products cited in the last two points are efforts by
> many other groups. We just pool them together to be EzIP building blocks.
>
>     Hope the above clears the air.
>
> Regards,
>
>
> Abe (2025-05-06 23:14 EDT)
>
>
>
> On 2025-05-06 12:17, Tom Beecher wrote:
>
> Suriya-
>
> Just so it is clear, the technology that Abraham is referencing ( EZIP )
> is ONLY a proposal that has been made. It has not been accepted by any
> standards body. It is not implemented or supported by any major router
> vendor. It will not work for you.
>
>
> On Tue, May 6, 2025 at 11:41 AM Abraham Y. Chen via NANOG <
> nanog () lists nanog org> wrote:
>
> > Hi, Suriya:
> >
> > 0)    I am glad that you requested off-list follow-ups, because what I
> > am going to share is quite controversial. With a general distribution
> > list, a discussion can easily be pulled off the track by personal /
> > emotional opinions or business interests, as you might have noticed on
> > the NANOG Forum in the past.
> >
> > 1)    I would recommend you to consider replacing 100.64/10 netblock
> > with 240/4 netblock for the CG-NAT configuration. This will reduce your
> > need for IPv4 addresses by 64 fold, thus mitigating the IPv4 address
> > shortage that you are facing.
> >
> > 2)    Although there have been (and still are) various attempts to make
> > use of the 240/4 netblock, none has approached it in a universal sense
> > as our proposal, called EzIP (phonetic for Easy IPv4). Others are either
> > piecemeal solutions for special cases or limited scope applications.
> > They will fragment the Internet and lead to chaos.  Characterized by
> > Vint Cerf as an "Overlay Network", EzIP scheme forms a new layer of
> > communication infrastructure that is independent of, yet in arm's-length
> > with the current Internet core. So that, EzIP can retain the desired
> > properties of the existing Internet, while shaking off the handicaps.
> > The former maintains the operation characteristics as CG-NAT to avoid
> > perturbing users, while the latter enables the Internet revamping into a
> > new era. This far-reaching implication is possible because EzIP resolves
> > the most fundamental issue of user identification resources. From such,
> > many constraints are either relaxed or simply removed.
> >
> > 3)    For a general introduction, please have a look at the below pair
> > of documents.
> >
> > https://avinta.com/gallery/DeterministicInternetIntro-US.pdf
> >
> > https://avinta.com/gallery/DeterministicInternet-SPKR.pdf
> >
> > 4)    Since this topic touches many aspects of the Internet and we are
> > not an operatorbut just a system analyst, we likely have not covered
> > many aspects that hands-on parties like you are familiar with. Please
> > browse through our website to see other background information which may
> > be relevant, then let us know your concerns. So that we can evaluate
> > them for you.
> >
> > Regards,
> >
> >
> > Abe (2025-05-06 11:40 EDT)
> > VP Engineering
> > Avinta Communications, Inc.
> > Milpitas, CA 95035 USA
> > O: +1(408)942-1485x66
> > M: +1(650)248-1829
> > Teams: Abraham.Y.Chen
> > eMail: AYChen () Avinta com
> > WebSite: www.Avinta.com
> >
> >
> > On 2025-05-06 04:36, Suriya Kamon via NANOG wrote:
> > > Hi NANOG,
> > >
> > > We are running short of IPv4 addresses.
> > >
> > > We are a small ISP and longer prefixes are okay with us (even /24s).
> > >
> > > Please contact me off-list.
> > >
> > > (Proper ROA coverage is a must).
> > >
> > > Thanks.
> > >
> > > Best Regards,
> > > Suriya
> > > _______________________________________________
> > > NANOG mailing list
> > https://lists.nanog.org/archives/list/nanog () lists nanog org/message/3MYN3WEGPZEMF44R2X2EG3UO7SAAHXMZ/
> >
> >
> >
> > --
> > This email has been checked for viruses by Avast antivirus software.
> > www.avast.com
> > _______________________________________________
> > NANOG mailing list
> >
> > https://lists.nanog.org/archives/list/nanog () lists nanog org/message/LTKZLWUSLFOSYEP5PHDXKMNWKZKOZ63V/
>
>
>
>
> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>
> Virus-free.www.avast.com
> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>
> <#m_4161442180352669907_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
_______________________________________________
NANOG mailing list 
https://lists.nanog.org/archives/list/nanog () lists nanog org/message/G6QHPB6MKVHR4ZKLD6JDFHUP43H652MA/