Re: [v6ops] Re: [Witarea] Re: How to make an elegant IPv4 outage

[Michael Richardson <[e-mail]>]

Phillip Hallam-Baker <[e-mail]> wrote:
    > On Fri, Jun 12, 2026 at 5:32 AM Tim Chown <[e-mail]> wrote:

    >> Perhaps we need to start draft-v6ops-ipv6-dinosaurs and start to name
    >> and shame such organisations and what they’re doing wrong… polite but
    >> objective critique and pointers in the doc to best practice.
    >>
    >> Tim
    >>

    > No, we need the exact opposite. We need branding people can recognize
    > as meaning that an ISP is providing an Internet service that is fit for
    > a particular purpose.

Yes, it needs to be branding/marketing, and it's not something that IETF does
well, or for which an RFC is particularly useful.

    > So we do an RFC that sets out distinct levels of service with some
    > indication of currency.

I would try to get a conclave of RIRs to do BCOPs on this, perhaps
coordinated via ICANN.  (But, really, ICANN itself should not be involved).
Maybe CableLabs / Broadband Forum are better venues.

    > INET-M-26: Mobile service, provides single dynamic IPv6 plus IPv4 which
    > may be behind a wide-cone NAT.

    > INET-R-26: Residential service, provides dynamic IPv6/56 plus an IPv4
    > connection that may be behind a wide-cone NAT

    > INET-S-26: Standard hosting service, provides one or more static IPv4
    > addresses plus a static IPv6/56.

Please note that there are two axes here:
1) if your v6 prefix is random/changeable vs deterministic.
2) if your v6 prefix is delivered by DHCPv6 or manually configured.

This is a detail, but it does affect ability to pick up and move ISP, while
retaining the same router.
"Standard hosting service" is the wrong term, and will be confusing to ISP
marketing types.   Because it's not about public facing web sites, even if
it probably involves port 443.

It should be "Small Business with remote security needs" ("hairdresser")

    > INET-H-26: High availability hosting service, routes IPv4 and IPv6
    > traffic to an ASN held by the end customer.

ISPs that can provide this already know what this is, it need only be defined
so that an ISP can say to the customer, who should be asking for "INET-S-26",
if that's really what they wanted.

Having said this, I've dealt with many L1/L2 providers who do not listen to
me when I explain, slowly, and clearly, that I've asked for L3 services,
and that's okay if they don't have that.  I've also regularly had annoying
attempts to upsell me to L3, when I really really wanted L2.

--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-                      *I*LIKE*TRAINS*