From: Joe Abley <jabley@hopcount.ca>
Message-Id: <C45C4CF8-FD43-49B2-B273-FAEDA05885F6@hopcount.ca>
Content-Type: multipart/signed;
 boundary="Apple-Mail=_300B4B4E-180D-4FC0-8B13-4AE342893ABB";
 protocol="application/pgp-signature"; micalg=pgp-sha1
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.80.23.2.2\))
Date: Thu, 30 Apr 2020 20:07:12 -0400
In-Reply-To: <8F265AC8-9369-40B6-9AE8-C8D8ED190320@isc.org>
Cc: Wes Hardaker <wjhns1@hardakers.net>, Tim Wicinski <tjw.ietf@gmail.com>,
 dnsop <dnsop@ietf.org>, dnsop-chairs <dnsop-chairs@ietf.org>
To: Mark Andrews <marka@isc.org>
References: <CADyWQ+FLrTy0gy8iCyAPsDpiumDNQHX4TGPni43ThA=W3fmZew@mail.gmail.com>
 <EB400743-8B25-45DA-B4BD-5B27F47AE9E3@hopcount.ca>
 <ybl5zdg4po9.fsf@w7.hardakers.net>
 <7262A449-1171-49E8-BDF6-69601DB034EE@hopcount.ca>
 <yblr1w438fb.fsf@w7.hardakers.net>
 <8F265AC8-9369-40B6-9AE8-C8D8ED190320@isc.org>
Archived-At: <https://mailarchive.ietf.org/arch/msg/dnsop/Gx0yrncpkvZ7_ZYs1ImEYaKvpGI>
Subject: Re: [DNSOP] Call for Adoption: draft-pwouters-powerbind
Precedence: list


--Apple-Mail=_300B4B4E-180D-4FC0-8B13-4AE342893ABB
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=utf-8

Hi Mark,

On 30 Apr 2020, at 19:52, Mark Andrews <marka@isc.org> wrote:

> On 1 May 2020, at 08:39, Wes Hardaker <wjhns1@hardakers.net> wrote:

>=20
>> Yep, I suspect some of the bigger TLDs probably couldn't opt in to =
this
>> draft simply because they're full of, um, "history".  Until that =
history
>> is cleaned, they probably couldn't deploy it.
>=20
> I doubt there is any real history here other than =E2=80=9Cwe didn=E2=80=
=99t properly clean
> up the records when the delegation was removed=E2=80=9D.  If a DNS =
provider goes belly
> up then all the zones dependent on that provider should FAIL.  Its =
just poor
> zone hygiene that leaves orphaned glue in the ORG zone.

I don't want to speculate on the actual reasons that orphan glue exists =
in the ORG zone before actually taking the time to investigate, but I'll =
note that the EPP data model provides some constraints around host =
objects and domain objects (e.g. domain objects can't be deleted when =
there subordinate host objects that exist) and there are reasons for =
delegations to be removed from the TLD zone while the corresponding =
domain object in the registry still exists (e.g. as part of domains that =
are suspended through the normal processes by registrars when they are =
found to be abusive).

I also don't understand the theory that any of these glue records exist =
because a DNS provider ceases operations. DNS providers don't feature in =
the RRR model. A failure in a set of nameservers could cause delegations =
to become lame, but there is no process that I'm aware of where orphan =
glue is a direct or necessary consequence.

Anyway, I am fairly confident in saying that there are legitimate, =
normal operational processes that can result in orphan glue, and that =
it's not correct to infer that they all exist for reasons of poor =
hygiene.

This is a topic of interest in a number of places right now. It is =
definitely on the list of things to investigate.


Joe

--Apple-Mail=_300B4B4E-180D-4FC0-8B13-4AE342893ABB
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename=signature.asc
Content-Type: application/pgp-signature;
	name=signature.asc
Content-Description: Message signed with OpenPGP

-----BEGIN PGP SIGNATURE-----

iF0EARECAB0WIQSAt40QkiztAcvphdg0jwy9hlI6LAUCXqtoMAAKCRA0jwy9hlI6
LD35AKCn8mLh243xRBZd0MOodvFsjMvuVQCgzFNegi/x+2XUD4cELgWJEaPDv8E=
=bt6e
-----END PGP SIGNATURE-----

--Apple-Mail=_300B4B4E-180D-4FC0-8B13-4AE342893ABB--