From owner-namedroppers@ops.ietf.org  Sat Sep  1 12:48:42 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA16031
	for <dnsext-archive@lists.ietf.org>; Sat, 1 Sep 2001 12:48:40 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15dDUT-000MT7-00
	for namedroppers-data@psg.com; Sat, 01 Sep 2001 09:19:25 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15dDUT-000MT1-00
	for namedroppers@ops.ietf.org; Sat, 01 Sep 2001 09:19:25 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15dDUS-0004oM-00
	for namedroppers@ops.ietf.org; Sat, 01 Sep 2001 09:19:24 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: namedroppers@ops.ietf.org
Subject: list policy
Message-Id: <E15d7ZI-000GKc-00@psg.com>
From: Randy Bush <randy@psg.com>
Date: Sat, 01 Sep 2001 03:00:00 -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

namedroppers@ops.ietf.org is the mailing list for the ietf dnsext wg.  see
<http://www.ietf.org/html.charters/dnsext-charter.html> for the wg charter.
messages should be on topics appropriate to the dnsext wg, which are various
discussion of the dns protocols or administrivia of the wg itself.  the list
is moderated.

calls for papers, announcements of events not directly relevant to the dns
protocols, etc. are not appropriate.  discussion of problems with particular
implementations, announcements of releases, sites' misconfigurations, etc.
should be done on mailing lists for the particular implementations.

there is a wg for dns operational practice, dnsop, whose charter can be
found at <http://www.ietf.org/html.charters/dnsop-charter.html>.

there is a mailing list for discussion of whose implementation is better,
and why someone else's is broken.  it is weenie-war@ops.ietf.org.  all
discussions of such nature should occur there or on /dev/null.  unlike the
namedroppers list, weenie-war@ops.ietf.org is not archived.

the general topic of mailing list moderation is a topic for the poisson wg.
poisson's charter can be found at
   <http://www.ietf.org/html.charters/poisson-charter.html>
questions or concerns related to the acceptance or rejection of specific
messages to the namedroppers mailing list should first be discussed with the
wg chairs, with followup appeals using the normal appeals process of rfc
2026 (i.e, followup with area directors, then iesg, etc.).

---

NOTE WELL:

All statements related to the activities of the IETF and addressed to the 
IETF are subject to all provisions of Section 10 of RFC 2026, which grants 
to the IETF and its participants certain licenses and rights in such 
statements.

Such statements include verbal statements in IETF meetings, as well as 
written and electronic communications made at any time or place, which are 
addressed to

    - the IETF plenary session,
    - any IETF working group or portion thereof,
    - the IESG, or any member thereof on behalf of the IESG,
    - the IAB or any member thereof on behalf of the IAB,
    - any IETF mailing list, including the IETF list itself,
      any working group or design team list, or any other list
      functioning under IETF auspices,
    - the RFC Editor or the Internet-Drafts function

Statements made outside of an IETF meeting, mailing list or other function, 
that are clearly not intended to be input to an IETF activity, group or 
function, are not subject to these provisions.



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep  4 12:36:59 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA20747
	for <dnsext-archive@lists.ietf.org>; Tue, 4 Sep 2001 12:36:58 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15eIlW-000Hzz-00
	for namedroppers-data@psg.com; Tue, 04 Sep 2001 09:09:30 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15eIlW-000Hzt-00
	for namedroppers@ops.ietf.org; Tue, 04 Sep 2001 09:09:30 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15eIlW-0001k7-00
	for namedroppers@ops.ietf.org; Tue, 04 Sep 2001 09:09:30 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109041021.f84AL5O33110@open.nlnetlabs.nl>
Subject: how to handle DS?
To: namedroppers@ops.ietf.org
Date: Tue, 4 Sep 2001 12:21:05 +0200 (CEST)
CC: team@open.nlnetlabs.nl
From: Alexis Yushin <alexis@NLnetLabs.nl>
Reply-To: alexis@NLnetLabs.nl
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Hi,

We're currently looking at the possibility of implementing DS
records for BIND 9 here at NLnet Labs... I could not find any
information on how DS records should be handled when actual
resolving takes place. Below is my idea of how it should work.

Imagine we have a secured root and a cache forwarder trying to
resolve the following query:

Q1 .		www.a.b.	A

A1 . -aa

	answ:	-

	auth:	b.		NS	ns.b.
		b.		DS .
		b.		SIG DS .

	add:	.		KEY
		.		SIG KEY
		ns.b.		A glue

Q2 ns.b.	www.a.b.	A

A2 ns.b. -aa

	answ:	-

	auth:	a.b.		NS	ns.a.b.
		a.b.		DS
		a.b.		SIG DS

	add:	b.		KEY
		b.		SIG KEY
		ns.a.b.		A	ip-address

Q3 ns.a.b.	www.a.b.	A

A3 ns.a.b. +aa

	answ:	www.a.b.	A	ip-address
		www.a.b.	SIG	A ip-address

	auth:	a.b.		NS	ns.a.b
		a.b.		SIG	NS

	add:	a.b.		KEY
		a.b.		SIG KEY
		ns.a.b.		A	ip-address


Is this correct? Any comments?

Alexis


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep  5 00:27:32 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA05572
	for <dnsext-archive@lists.ietf.org>; Wed, 5 Sep 2001 00:27:31 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15eTuh-000Bez-00
	for namedroppers-data@psg.com; Tue, 04 Sep 2001 21:03:43 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15eTuh-000Bet-00
	for namedroppers@ops.ietf.org; Tue, 04 Sep 2001 21:03:43 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15eTuh-000L4U-00
	for namedroppers@ops.ietf.org; Tue, 04 Sep 2001 21:03:43 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <5.1.0.14.2.20010904230246.023472b0@localhost>
Date: Tue, 04 Sep 2001 23:10:48 -0400
To: alexis@NLnetLabs.nl, namedroppers@ops.ietf.org
From: =?iso-8859-1?Q?=D3lafur?= =?iso-8859-1?Q?_Gu=F0mundsson?= <ogud@ogud.com>
Subject: Re: how to handle DS?
Cc: team@open.nlnetlabs.nl
In-Reply-To: <200109041021.f84AL5O33110@open.nlnetlabs.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

At 06:21 AM 9/4/2001, Alexis Yushin wrote:
>Hi,
>
>We're currently looking at the possibility of implementing DS
>records for BIND 9 here at NLnet Labs... I could not find any
>information on how DS records should be handled when actual
>resolving takes place. Below is my idea of how it should work.
>
>Imagine we have a secured root and a cache forwarder trying to
>resolve the following query:
>
>Q1 .            www.a.b.        A
>
>A1 . -aa
>
>         answ:   -
>
>         auth:   b.              NS      ns.b.
>                 b.              DS .
>                 b.              SIG DS .
>
>         add:    .               KEY
>                 .               SIG KEY
>                 ns.b.           A glue

I have assumed that the DS would go in the additional section (unless it is
in the answer section).


>Q2 ns.b.        www.a.b.        A
>
>A2 ns.b. -aa
>
>         answ:   -
>
>         auth:   a.b.            NS      ns.a.b.
>                 a.b.            DS
>                 a.b.            SIG DS
>
>         add:    b.              KEY
>                 b.              SIG KEY
>                 ns.a.b.         A       ip-address

almost, A has higher priority than KEY thus it should be before the
Question should the DS have higher priority than the KEY for a. ?

>Q3 ns.a.b.      www.a.b.        A
>
>A3 ns.a.b. +aa
>
>         answ:   www.a.b.        A       ip-address
>                 www.a.b.        SIG     A ip-address
>
>         auth:   a.b.            NS      ns.a.b
>                 a.b.            SIG     NS
>
>         add:    a.b.            KEY
>                 a.b.            SIG KEY
>                 ns.a.b.         A       ip-address

Other than order in additional section this looks about right.

         Olafur



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep  5 01:24:20 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA05571
	for <dnsext-archive@lists.ietf.org>; Wed, 5 Sep 2001 00:27:31 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15eTuR-000Beh-00
	for namedroppers-data@psg.com; Tue, 04 Sep 2001 21:03:27 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15eTuR-000Bea-00
	for namedroppers@ops.ietf.org; Tue, 04 Sep 2001 21:03:27 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15eTuQ-000L42-00
	for namedroppers@ops.ietf.org; Tue, 04 Sep 2001 21:03:26 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <5.1.0.14.2.20010904231058.023499a0@localhost>
Date: Tue, 04 Sep 2001 23:21:48 -0400
To: DNSEXT <namedroppers@ops.ietf.org>
From: =?iso-8859-1?Q?=D3lafur?= =?iso-8859-1?Q?_Gu=F0mundsson?= <ogud@ogud.com>
Subject: Re: I-D ACTION:draft-schlyter-appkey-00.txt
In-Reply-To: <Pine.BSO.4.33.0108311103420.11836-100000@fonbella.crt.se>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

At 05:05 AM 8/31/2001, Jakob Schlyter wrote:

Please take a look at these two ID's both are trying to address the
the problem with sub typing of KEY records and the possibility of
large KEY RRset.


>         Title           : Storing application public keys in the DNS
>         Author(s)       : J. Schlyter
>         Filename        : draft-schlyter-appkey-00.txt
>         Pages           : 8
>         Date            : 29-Aug-01
>
>This document specifies a new DNS RR type for applications to store
>public keys in.  Experience with DNSSEC has indicated that mixing DNS
>keys and application keys is a bad idea in many regards.  The new RR
>expands certain fields based on experience from early DNSSEC
>deployment.  This document updates RFC 2535.
>
>A URL for this Internet-Draft is:
>http://www.ietf.org/internet-drafts/draft-schlyter-appkey-00.txt


>         Title           : DNS KEY Resource Record Generic Protocol Value
>         Author(s)       : E. Lewis
>         Filename        : draft-lewis-dnsext-key-genprot-00.txt
>         Pages           :
>         Date            : 10-Jul-01
>
>A new protocol value is defined for the KEY Resource Record which
>identifies the intended protocol as that identified in the SRV-like
>encoding of the KEY RR's owner name.
>
>A URL for this Internet-Draft is:
>http://www.ietf.org/internet-drafts/draft-lewis-dnsext-key-genprot-00.txt


The total solution space now is
0. Do nothing leave KEY records as is and not worry.
1. Generic Protocol value for KEY record
2. APPKEY record
3. Create a new KEY type record for every protocol that wants
    that wants to store


         Olafur



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep  5 10:59:20 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA06156
	for <dnsext-archive@lists.ietf.org>; Wed, 5 Sep 2001 10:59:20 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15edT4-000NeL-00
	for namedroppers-data@psg.com; Wed, 05 Sep 2001 07:15:50 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15edT3-000NeE-00
	for namedroppers@ops.ietf.org; Wed, 05 Sep 2001 07:15:49 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15edT3-000Brs-00
	for namedroppers@ops.ietf.org; Wed, 05 Sep 2001 07:15:49 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Date: Wed, 5 Sep 2001 10:20:08 +0200 (MEST)
From: Jakob Schlyter <jakob@crt.se>
To: =?iso-8859-1?Q?=D3lafur?= =?iso-8859-1?Q?_Gu=F0mundsson?= <ogud@ogud.com>
Cc: DNSEXT <namedroppers@ops.ietf.org>
Subject: Re: I-D ACTION:draft-schlyter-appkey-00.txt
In-Reply-To: <5.1.0.14.2.20010904231058.023499a0@localhost>
Message-ID: <Pine.BSO.4.33.0109051014520.15752-100000@fonbella.crt.se>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 8bit

On Tue, 4 Sep 2001, Ólafur Guðmundsson wrote:

> 0. Do nothing leave KEY records as is and not worry.
> 1. Generic Protocol value for KEY record
> 2. APPKEY record
> 3. Create a new KEY type record for every protocol that wants
>     that wants to store

and, even though I personally do not support it, we also have

4. use CERT record for all application keys/certificates.


some people have suggested that we add features of CERT to APPKEY and then
deprecated CERT leaving us with the infrastructure KEY type and the
application APPKEY type.


	jakob



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep  5 11:10:02 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA06775
	for <dnsext-archive@lists.ietf.org>; Wed, 5 Sep 2001 11:10:01 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15edq6-000OAe-00
	for namedroppers-data@psg.com; Wed, 05 Sep 2001 07:39:38 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15edq6-000OAY-00
	for namedroppers@ops.ietf.org; Wed, 05 Sep 2001 07:39:38 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15edq6-000CXX-00
	for namedroppers@ops.ietf.org; Wed, 05 Sep 2001 07:39:38 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Message-ID: <005501c13609$16f14240$b9370681@antd.nist.gov>
From: "Scott Rose" <scottr@antd.nist.gov>
To: "DNSEXT" <namedroppers@ops.ietf.org>
References: <5.1.0.14.2.20010904231058.023499a0@localhost>
Subject: Re: I-D ACTION:draft-schlyter-appkey-00.txt
Date: Wed, 5 Sep 2001 08:48:40 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 8bit

----- Original Message -----
From: "Ólafur Guðmundsson" <ogud@ogud.com>
To: "DNSEXT" <namedroppers@ops.ietf.org>
Sent: Tuesday, September 04, 2001 11:21 PM
Subject: Re: I-D ACTION:draft-schlyter-appkey-00.txt

> The total solution space now is
> 0. Do nothing leave KEY records as is and not worry.
> 1. Generic Protocol value for KEY record
> 2. APPKEY record
> 3. Create a new KEY type record for every protocol that wants
>     that wants to store

I think options 0 and 3 are not the front runners.  I haven't heard an
arguement for leaving the KEY record as it stands, and allocating a new type
for every protocol is just another hurdle for protcol developers.  So that
leaves options 1 and 2 and the compromise Håvard suggested (my
interpretation):  The CERT record SHOULD be used when the protocol uses a
public key and a non-DNS signature and performs verification/authentication
out of band.  An APPKEY (or some other new solution, but APPKEY is
reasonable) SHOULD be used when the application relies on using the DNS
resolver to perform authentication.  (APPKEY and a SIG record).

There are other outstand issues as well, but focusing on one at a time is a
good idea for now.

Scott

to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep  5 11:16:04 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA06976
	for <dnsext-archive@lists.ietf.org>; Wed, 5 Sep 2001 11:16:02 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ee0D-000OTA-00
	for namedroppers-data@psg.com; Wed, 05 Sep 2001 07:50:05 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ee0D-000OT4-00
	for namedroppers@ops.ietf.org; Wed, 05 Sep 2001 07:50:05 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ee0D-000Coz-00
	for namedroppers@ops.ietf.org; Wed, 05 Sep 2001 07:50:05 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <v0313030bb7bbd81a2e2b@[208.58.215.115]>
In-Reply-To: <5.1.0.14.2.20010904231058.023499a0@localhost>
References: <Pine.BSO.4.33.0108311103420.11836-100000@fonbella.crt.se>
Date: Wed, 5 Sep 2001 09:26:15 -0400
To: =?iso-8859-1?Q?=D3lafur?= =?iso-8859-1?Q?_?= =?iso-8859-1?Q?Gu=F0mundsson?=  <ogud@ogud.com>
From: Edward Lewis <lewis@tislabs.com>
Subject: Re: I-D ACTION:draft-schlyter-appkey-00.txt
Cc: DNSEXT <namedroppers@ops.ietf.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Another is create a CERT RR certificate type to hold public keys.  This has
been suggested on the dnssec@cafax.se but is not documented.  (I know -
send text, but this isn't my idea, nor one I have been supporting.  But I
felt it should be added to the list here.)

At 11:21 PM -0400 9/4/01, =D3lafur Gu=F0mundsson wrote:
>At 05:05 AM 8/31/2001, Jakob Schlyter wrote:
>
>Please take a look at these two ID's both are trying to address the
>the problem with sub typing of KEY records and the possibility of
>large KEY RRset.
>
>
>>         Title           : Storing application public keys in the DNS
>>         Author(s)       : J. Schlyter
>>         Filename        : draft-schlyter-appkey-00.txt
>>         Pages           : 8
>>         Date            : 29-Aug-01
>>
>>This document specifies a new DNS RR type for applications to store
>>public keys in.  Experience with DNSSEC has indicated that mixing DNS
>>keys and application keys is a bad idea in many regards.  The new RR
>>expands certain fields based on experience from early DNSSEC
>>deployment.  This document updates RFC 2535.
>>
>>A URL for this Internet-Draft is:
>>http://www.ietf.org/internet-drafts/draft-schlyter-appkey-00.txt
>
>
>>         Title           : DNS KEY Resource Record Generic Protocol Value
>>         Author(s)       : E. Lewis
>>         Filename        : draft-lewis-dnsext-key-genprot-00.txt
>>         Pages           :
>>         Date            : 10-Jul-01
>>
>>A new protocol value is defined for the KEY Resource Record which
>>identifies the intended protocol as that identified in the SRV-like
>>encoding of the KEY RR's owner name.
>>
>>A URL for this Internet-Draft is:
>>http://www.ietf.org/internet-drafts/draft-lewis-dnsext-key-genprot-00.txt
>
>
>The total solution space now is
>0. Do nothing leave KEY records as is and not worry.
>1. Generic Protocol value for KEY record
>2. APPKEY record
>3. Create a new KEY type record for every protocol that wants
>    that wants to store
>
>
>         Olafur
>
>
>
>to unsubscribe send a message to namedroppers-request@ops.ietf.org with
>the word 'unsubscribe' in a single line as the message text body.


-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D--=3D-=
=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-=3D-
Edward Lewis                                                NAI Labs
Phone: +1 443-259-2352                      Email: lewis@tislabs.com

You fly too often when ... the airport taxi is on speed-dial.

Opinions expressed are property of my evil twin, not my employer.




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 10 00:52:13 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA20937
	for <dnsext-archive@lists.ietf.org>; Mon, 10 Sep 2001 00:52:12 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15gI7j-000KJu-00
	for namedroppers-data@psg.com; Sun, 09 Sep 2001 20:52:39 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15gI7j-000KJo-00
	for namedroppers@ops.ietf.org; Sun, 09 Sep 2001 20:52:39 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15gI7j-000P2L-00
	for namedroppers@ops.ietf.org; Sun, 09 Sep 2001 20:52:39 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <5.1.0.14.2.20010909175154.00b2cb10@localhost>
Date: Sun, 09 Sep 2001 22:54:17 -0400
To: namedroppers@ops.ietf.org
From: Olafur Gudmundsson <ogud@ogud.com> (by way of =?iso-8859-1?Q?=D3lafur?=
 =?iso-8859-1?Q?_Gu=F0mundsson?= <ogud@ogud.com>)
Subject: IETF 51, DNSEXT final minutes
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

DNSEXT IETF-51 meeting 2001/August/9 9:00-11:30 GMT.  Final minutes
Chairs:	       Randy Bush
	       Olafur Gudmundsson


WG STATUS:                         Olafur Gudmundsson
- 1 new RFC: APL
- 2 documents at RFC editor, message size + DNSSEC OK bit
- 2 docs at IESG: DNSSEC road map + RFC161{1,2} retire
- AXFR has passed working group call
- MUST figure out DNSSEC status
- Please adopt RFC to advance

Summary of joint DNSEXT/NGTRANS meeting       Olafur Gudmundsson
- DNSEXT will take lead on standardizing IPv6 address related RFCs.
- A6, bitlabels => experimental
- DNAME standardization is by DNSEXT it stays for now at proposed.
- AAAA to go to drafter standard, need volunteer to adopt for interop
   testing and reporting.

The main focus of this meeting was to asses the current DNSSEC status
and how it should evolve from now on. First there is a presentation on
the status and history of DNSSEC, followed by presentations about new
proposals and concluding by an open discussion on how to go forward.

Goal: determine how close to operationally deployable DNSSEC
Subgoal: collect list of threats to DNS that only DNSSEC can address
Output: proposed plan of action posted to mailing list

- ---
Current DNSSEC status and options on how to progress
                                         Edward Lewis + Olafur Gudmundsson

context
DNSSEC had a spurt of definition from 1994 to 1998
partial implementation in bind822 1999
workshop followed
	bind9 full 2535 DNSSEC with tweaks
	Jnamed has almost full DNSSEC compliance
recently, as a result of experiments
	an explosion in issues, discussion, activity

set forth a few options regarding to the future of DNSSEC
judge the severity and volume of current issues to see what outcome is
recommended no intention to discuss the issues or solutions here

DNSSEC seems to be healthy in terms of interest but with its potential
importance and this healthy interest, why aren't we done yet?

what could be done:
  - allow just some tweaking
  - pursue significant changes
  - scrap the current approach and start over
  - freeze the current design and advance it

allow just some tweaking:
means leaving most of the definition intact
can we solve the rest of the problems with tweaks and adjustments?
or are we running in circles before it all come crashing down?

major surgery:
a few IETFs ago (oslo or dc) an edict was set forth that DNSSEC must
settle down too many theorists, not enough operators
the hope was to give the current definition a real litmus test
how much longer to we wait for the test to begin? or should the edict
be rescinded?  has the test begun already or not?

scrap the current approach and start over
well, at least we've learned a lot
perhaps the initial conditions set forth are doomed to lead us to a dead end
is dns1.0 (rfc1035) inherently unsecurable?
perhaps we need to restart, perhaps with dns2.0
	threat model, requirement document are needed

freeze the current design and advance it


issues
(somewhat arbitrary division of the issues)
	CERT records
	TSIG operations
		management, TKEY
	SIG/KEY operations
		exposure of data, delegation management, dynamic update
	applications and stub resolution
	authority at delegations

CERT records
nothing to mention really
not much activity, but this isn't a fault of dns
one issue area down...
	just a doc, noone is using it.

TSIG
- Workshops demonstrated that TSIG is maturing well.
- Work has lagged in secret management.
- Two pending drafts, GSS-API and TKEY renewal.
- Open issue is use of TSIG in general purpose queries.

SIG/KEY
a much more problematic area
basic operations are defined, but not always scalable
some issues are even recognized as issues
   - 	exposure of data - via authenticated denial
other issues
   -	delegation management
   -	dynamic update refresh

SIG/KEY exposure
NXT record
	different approach: NO
	modified approach - allowing exclusion of sets from NXT
this issue relates to delegation management
	excluding records from NXT eases burden
	provides cheaper means to signify unsecured delegations

exposure drafts
	opt-in-00, rrsets-00, not-existing-rr-00

delegation management
difficult set of issues
	RFC1035 has the same delegation records in parent and child,
	with the child one more credible, and no record type that only
	exist in parent
DNSSEC requires more complex communication between parent and child
	registrar, registry, registrant complicates
KEY signing opens up
	liability issues for parents
	QoS issues for child

SIG/KEY at delegations
APEX KEY set can be large, change frequently contain non DNS zone KEY records
proposals to solve problems
	reduce communication between parent and child
	allow partial security in secure zones
	allow zones to express security policies in use

	parent-store-zone-keys-01, delegation-signer-01,
	dnssec-opt-in-00, dnsext-rrsets-00, sec-rr-00

SIG/KEY dynamic update
this topic suffers from a lack of activity
early experiments noted problems in early implementations
	may have been fixed
	but no one is exercising this area (publicly)

	in order to sign zone, you have to have private key present on
	the server

application and stubs
how are answers to be interpreted? what are the meaning of the bits in
the header?
should application keys be stored differently than DNSSEC keys?
	dnssec-okbit02, key-genprot-00, ad-is-secure-03

design tradeoffs
delegation maintenance is hard, thus some new proposals suggest
changes that increase resolver work and add latency in protocol
is it acceptable to secure only fraction of a zone?
when there is choice what part of operation should be favored,
operation or resolution?

anyone address some of the problems?

ohta: public key crypto violates e2e principle.

- ---
DNSSEC editing status            Dan Massey, Scott Rose
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-dnssec-intro-00.txt>
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-rfc2536bis-dsa-00.txt>
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-rfc2539bis-dhk-00.txt>

- ---
Opt IN plan                      Mark Kosters
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-dnssec-opt-in-00.txt>

opt-in for large zones
DNSSEC is painful for large zones
huge up front cost
	null keys, nxt rrs for unsecured child zones
little initial buy-in
high performance penalty on lookup (hash vs tree)
high generation cost
	ordering, signing, distribution
desire to have autonomous DNSSEC processing

changes since last version of draft.
dropped additional rcode "retry-no-sec" in response to DNSSEC queries
for delegations that are not secure
now answers from "unsecure" side folded in with secure response for a
query on a unsecured or non-existent domain so no need for query
semantics of NXT changed from "non-existence" to "secure but not in
range" indicated by setting bit 3 in KEY RR

pros
- incremental cost to DNSSEC adoption
- independent processing for DNSSEC
- no changes for 1035 processing
- DNSSEC not widely deployed yet so now is the time for change
cons
- resolvers need to be updated to deal with opt-in (easy according to nominum)

pilot
implemented an opt-in pilot for com/net/org
     demonstrate ease of use, implementation for conformance testing
     use parent sig over child (for now)
     code is *not* based on bind
     3 example zones with web-based zone editor to show how easy it
     could be to secure/unsecure a zone with opt-in
     key mgmt web interface for other domains under com/net/org
http://www.research.netsol.com/

Q: updating recursive resolvers is hard.
more elegant solution needed than verisign needs to do.

Rob Austein: signing time issue is not quite it looks like,
    because it can be done in parallel, except when changing keys
    frequently, which you're not planning on doing.)

look at it at first

NO SIG                           Mark Kosters for Roy Arens
diff between NOsigRR and opt-in
  - NXT changed from non-existence to not-secure
  - OPT-IN defines this in KEY RR
  - NOSIG defines this with pseudo-RR hel in NXT type bitmap
  - Pros: accomplishes opt-in goals, adds in-zone RRs that could be unsecured,
   allows thin layer of security from zone to zone

Ed Lewis: Create a type different from NXT rather than trying to change
    its existing definition.


SIG@Parent                       Miek Gieben
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-parent-stores-zone-keys-01.txt>

good points:
   works very well operationally
   reduces administrator involvement
   makes key rollover possible

but:
   resolver must be adapted
   local (ssh) keys are also signed

solution:
   new record at the parent -> DS?

Wrote a tool to verify signature chains
   <http://www.nlnetlabs.nl/downloads/dig-9.2.0b2-nlnetlabs-0.1.tar.gz>


- ---
Delegation Signer                Olafur Gudmundsson
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-delegation-signer-01.txt>

APEX records issue.

motivations
key record repeats the mistake of NS record in delegations, same
record in two zones
key at child is cumbersome
key@parent has
	liability issues for parent
	QoS child is totally at parent mercy on changes
	parent may only allow dns keys in key set
is there another way?

DS advantages
  - minimizes communication
	only when child changes key set signed key
  - less data in parent
	DS much smaller than key
	DS only at secure delegations
	child in full control of security status
	no conflicting key sets

DS disadvantages
   - More work for resolves, worst case doubling number of sig
     verifications
   - Has to calculate checksums for keys


DS record
checksum - SHA1(canonical fqdn + key rdata)
data format: 25 bytes, fixed size
	key id (2), key size (2), algorithm, sha-1 digest
possible to make the record smaller.

Donald Eastlake commented that the statement in presentation about
childs independence was to strong.
Ian Jackson liked the flexibility this gives child for key policies.
Dan Massey: wondered about the right actions to take if child key gets
compromised, he will post discussion points on mailing list.

================================================================
SEC RR, Ed Lewis
Currently a place holder for discussion. IF and only if some of the new
ideas discussed such as opt-in, DS etc are adopted, it might be nice
to tell resolvers what to expect in each zone. SEC RR is indented
to express the policies and security records used.

================================================================
40 min Next steps in DNSSEC             Randy Bush
        This will be open mike session to air opinions on which of
        the following options to go with. The chairs will take the
        suggestions and results of straw polls to formulate a plan of
        action to propose to the working group on the the mailing list.
        The options are:
           -  Finished tweaking DNSSEC (no more changes)
           -  Few more tweaks  (which ones)
           -  Major Surgery (any collection new proposals such as
                         opt-in, delegation signer, no, no-sig,  )
           -  Start over, because the current approach is not going to work

	  -  Do we really need it?)

Tom Monday: List of issues is representative of what we know of today
but there are probably more we have not yet discovered.
Unfortunately the IETF is not good at establishing when things have
been explored enough.

Q: dynamic update part/stuff really works or not?
OG: Dynamic update needs more attention.
Randy: We need to have a serious, week long workshop with very few
people, but throwing in dynamic update with add at least another week,
maybe two.

Lars Liman: Cost of security is highly relevant, needs to be balanced
against the risk of losing money if you don't have it.  Need to ask
people how much they could lose without it, much they are willing to
invest in the hard work to protect the money they stand to lose?

Randy: How many people here seriously working on DNSSEC are funded by
commercial interests? [2.]  How many by military interests? [A few more.]

Rob Austein: Biggest single thing that keeps whacking us in the face
seems like we have a handle on it, but it would take a lot of time to
implement and test.  If we decided to undertake this, we need to go
back and re-examine the threat model and see just what problems we're
trying to address.  Think we could do dynamic update work in
parallel.  My major hope is the last major thing we really need to
change with DNSSEC is the delegation signing issue, so shouldn't run
into needs to major change the protocol further when work on dynamic
update is done.

Ian: A lot of problems are related to not having clear layering model,
that DNSSEC simultaneously looks like part of the normal, accepted DNS
layer, and part of it looks like a layer on top but with lots of
layering violations.

Donald: Wrote some of original documents but have not been all that
active recently.  The problem with multilevel hierarchal delegation is
intrinsically difficult with no working models anywhere.  Could move
the security communication completely out of the DNS -- but not clear
that would make a bit of difference.  If we have DNS security, we need
to know what services it is offering and focus on offering them
strongly.  Biggest mistake would be offering many services and
providing them with weak security.

Derek Atkins: One thing about security is that it is very pervasive,
it is going to violate layers when you add it to something that was
not designed with it, either that or you have to wholly replace the
system.  With regard to secure dynamic update, I did an implementation
in 1995 with the biggest issue being online vs offline signing.  You
could do it with delegating a signing key to the updater, like the
DHCP server.  In terms of the real problems that DNSSEC is trying to
solve, the biggest one is that DNSSEC solves the problem of name
referrals, getting from one name (a host abbreviation alias) to a full
name and hence its associated data.

Rip Loomis: Military spending a huge amount of money on trying to
make this work, consider it crucial.

Ed: SIG and KEY are where all the problems are, others seem to be
ready to go.

Michael Patton: Was agnostic about DNSSEC originally when chaired the
DNSng BOF, but now think it is very important to get something out there
ASAP.  We really need the experience of having a V1.0 in order to
learn enough to have a V2.0 which does it right.  In favor of doing
both very little tweaking to move forward, and also scrapping the
current approach to design from scratch a better oe.

Ian: We will always need to have a way to send secured data end-to-end
over unsecured path, with full service resolvers doing verification.

Olafur: Early adopters always face a risk.

Randy: I'm not a security weenie, but security experts I trust tell me
that complexity is inversely proportional to strength of security, and
having a complex process that is not terminating makes me really
uncomfortable about the ultimate security of this system.  Concerned
about the choice of pushing forward with the intent to fix things
later.  The DNS was bleedingly complex before we even tried to graft
DNSSEC into it.  Trying to figure out how to get comfortable with any
of the choices.

Russ Monday: From NAI, working with ISI on DNSSEC.  Important thing
NAI/ISI can do is provide a testbed for all this, willing to make
available to researchers from other organizations.  Sometimes one
person's tweak is another's significant change.  Should focus on the
delegation problem.  Should test various proposals and come back with
results from that for next IETF meeting.

Bill Manning: Part of the problem is that DNSSEC is a lot like porn in
that we claim to know it when we see it (implying there is not an
objective way to measure it).  ...missed... At some point we really do
need to start over with the design of it.  Also, I don't believe there
can be a single threat model, we can and should have multiple good
ones.

Derek Atkins: Regarding simplicity versus complexity, Einstein said,
"You should design your system to be as simple as possible and no
simpler," and I really think DNSSEC really is as simple as we can make it.

Rob: Propose parallel approach: work on threat model, and rigorously
test delegation proposals.

Sense of room seems to agree, no one spoke against.

Donald: Move TSIG to the side.

Ian: Can we have a definition on what's a tweak and what's a major change?

Randy: NO.

Matt Larson: Let's not forget the large zone problem!  Not as
significant as the large delegation zone problem, but still really
significant.

Mark Kosters: Since we have to change the recursive nameserver anyway,
we might as well address the large zone issue at the same time.
Have not liked NXT since the beginning, but I have become resigned to
authenticated denial.

Rob Austein: Confidentiality of data was explicitly NOT one of stated
goals of DNSSEC, so people who don't like NXT, tough noogies.  Also,
because of the interdependence of MX and A records, you can misroute
mail. Without authenticated denial there are other attacks possible.

Sense of the room is to pursue parallel work on:
  - delegation management
    Russ will pursue with NAI/ISI testbed
  - protocol discussion on partially signed (ie, large) zones
    Mark's work converging
  - get serious about dynamic update
    ... need someone to adopt, contact chairs
  - threat models / requirements
    Rob Austein and Derek Atkins

- ---
Old Working Group documents
05 min Multicast DNS                     Dave Thaler
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-mdns-03.txt>

- Use of "solicited name" multicast addresses for IPv6, using same
   mechanism as used elsewhere (icmp)
- Clarification of multihoming behavior
- dynamic update for conflict detection
- added .ARPA domain considerations section

Open Issues
- Configuration of mdns
   - still assumes domain search option for configuration, but there
   are other proposals.
- Draft useful as stands?
  - Original goal was to allow retirement of NetBIOS and AppleTalk in
  unconfigured networks.
  - Has this goal been met?

- ---
Relevant Documents in Other working groups
05 min DHC MulticastDNS enable option    Erik Guttman
    <http://www.ietf.org/internet-drafts/draft-guttman-dhc-mdns-enable-01.txt>

mdns enable dhc option
complete replacement to local.arpa zone

wants to replace appletalk with IP services.

================================================================

REPLACE APPLETALK & NETBIOS, Stuart Cheshire
draft-cheshire-dnsext-multicastdns-00

An alternative multicast DNS proposal.

Goals/features:
- Free choice of names for machines only on local link
- Name _services_ at first class entities, not just hosts
- Browse for services, like the printers on the network
- Failsafe reliability
- Simple to implement

================================================================

At this point working group was out of time, chair wanted to resolve
one issue, Olafur Gudmundsson asked Bill Manning a question:
If the working group wants to adopt your discover draft will you
change the copyright to boilerplate #1?
Bill Manning: YES
End of meeting: Number of agenda items did not get a hearing this time
including following documents.

- ---
New documents requesting to become working group documents
05 min TKEY rekey mode                   Yuji Kamite
    <http://www.ietf.org/internet-drafts/draft-kamite-dnsext-tkey-secret-renewal-00.txt>

- ---
05 min Generic KEY RR Protocol value     Edward Lewis
    <http://www.ietf.org/internet-drafts/draft-lewis-dnsext-key-genprot-00.txt>

- ---
05 min ECC KEY rr                        Donald Eastlake
    <http://www.ietf.org/internet-drafts/draft-ietf-dnsext-ecc-key-00.txt>



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 12 11:26:05 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA24660
	for <dnsext-archive@lists.ietf.org>; Wed, 12 Sep 2001 11:26:04 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hBB0-0001d3-00
	for namedroppers-data@psg.com; Wed, 12 Sep 2001 07:39:42 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hBAz-0001cx-00
	for namedroppers@ops.ietf.org; Wed, 12 Sep 2001 07:39:41 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hBAy-0001LE-00
	for namedroppers@ops.ietf.org; Wed, 12 Sep 2001 10:39:40 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Tue, 11 Sep 2001 22:30:24 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: namedroppers@ops.ietf.org
cc: aboba@internaut.com
Subject: mDNS configuration discussion summary
Message-ID: <Pine.BSF.4.21.0109112214010.58337-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

In the recent thread relating to mDNS configuration, concern has been
expressed about the overloading of the DNS searchlist for purposes of mDNS
configuration. As a result, several alternative proposals have been
advanced, including an mDNS enable search option, and extension of RFC
2937, the name service search list. 

Each of the proposals under discussion enables stateful configuration of
mDNS via a DHCP option. However, it was pointed out that only the DNS
searchlist can currently be discovered as part of the stateless DNS
Configuration Discovery (DDNS) proposals under development within
IPNG WG. 

In examining the RFC 2937 approach, several issues have arisen:

a. RFC 2937 currently uses DHCP option numbers in order to identify the
   name services in the searchlist. 

b. RFC 2937 only addresses the issue of search ordering; it does not
   provide a mechanism for enabling or disabling individual name
   resolution mechanisms. For example, if DNS is not included in the
   searchlist, this does not imply that DNS is not to be used,
   particularly if DNS server addresses are provided in another
   option. 

In order to address these issues, as well as to take into account the
concerns expressed in the thread, I would like to propose the following
approach:


1. Delete references to use of the DNS searchlist for purposes of
   mDNS configuration. Other than the issue of stateless discovery,
   (addressed below) there was not much support for use of this approach. 

2. Utilize the mDNS enable option in order to enable or disable mDNS.

3. Utilize RFC 2937 (using the mDNS option code allocated for the above
   option) to specify the order in which mDNS will be used relative to
   other name resolution mechanisms.  

4. Add support for stateless discovery of mDNS enable/disable to the DDNS
   proposals in progress in IPNG WG. Since the proposals enable discovery
   of attributes, this should not be difficult. 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 12 11:47:01 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA25422
	for <dnsext-archive@lists.ietf.org>; Wed, 12 Sep 2001 11:47:00 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hBiP-0002J3-00
	for namedroppers-data@psg.com; Wed, 12 Sep 2001 08:14:13 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hBiO-0002Ix-00
	for namedroppers@ops.ietf.org; Wed, 12 Sep 2001 08:14:12 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hBiN-0001Nn-00
	for namedroppers@ops.ietf.org; Wed, 12 Sep 2001 11:14:11 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 12 Sep 2001 17:22:07 +0200 (MET DST)
From: Erik Guttman <Erik.Guttman@Sun.COM>
Reply-To: Erik Guttman <Erik.Guttman@Sun.COM>
Subject: Re: mDNS configuration discussion summary
To: Bernard Aboba <aboba@internaut.com>
Cc: namedroppers@ops.ietf.org
In-Reply-To: "Your message with ID" <Pine.BSF.4.21.0109112214010.58337-100000@internaut.com>
Message-ID: <Roam.SIMC.2.0.6.1000308127.12342.erikg@ehdb03-home.germany>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> I would like to propose the following approach:
> 
> 1. Delete references to use of the DNS searchlist for...mDNS configuration.
>
> 2. Utilize the mDNS enable option in order to enable or disable mDNS.
> 
> 3. Utilize RFC 2937 to specify the order in which mDNS will be used 
> 
> 4. Add...stateless discovery of mDNS enable/disable to...DDNS...in...IPNG

I support this proposal.

Erik



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 13 14:27:53 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA12496
	for <dnsext-archive@lists.ietf.org>; Thu, 13 Sep 2001 14:27:52 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hacc-000Fj3-00
	for namedroppers-data@psg.com; Thu, 13 Sep 2001 10:49:54 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hacb-000Fix-00
	for namedroppers@ops.ietf.org; Thu, 13 Sep 2001 10:49:53 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15haca-0002ma-00
	for namedroppers@ops.ietf.org; Thu, 13 Sep 2001 13:49:52 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 13 Sep 2001 09:23:21 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: namedroppers@ops.ietf.org
Subject: New mDNS configuration language
Message-ID: <Pine.BSF.4.21.0109130920090.1339-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

I would like to propose the following text on mDNS configuration be added
to the next version of the draft, and that previous text relating to use
of the domain searchlist for configuration be removed. Comments welcome. 

3.1.  mDNS configuration

Multicast DNS usage can be configured manually or automatically.  Where
no manual or automatic configuration is provided, multicast DNS is
enabled by default.

For IPv6, the stateless DNS discovery mechanisms described in [19] can
be used to discover whether multicast DNS is enabled or disabled.

Where DHCPv4 or DHCPv6 is implemented, DHCP options can be used to
configure multicast DNS. The mDNS enable DHCP option, described in [6],
can be used to explicitly enable or disable use of multicast DNS. The
Name Service Search option, described in RFC 2937 [3], can be used to
determine where multicast DNS is used within the name service search
order. DHCP option codes are used as RFC 2937 codes signifying name
services within the search order. As a result, to specify multicast DNS
usage within the name service search order, the option code assigned to
the mDNS enable option is used.

If a host is configured via automatic configuration mechanisms, either
stateful or stateless, and multicast DNS is not explicitly enabled, then
multicast DNS MUST NOT be used, ensuring that upgraded hosts do not
change their default behavior.  This implies that the host will neither
listen on the DNS LINKLOCAL multicast address, nor will it send queries
to that address.  For a DNS server, automatic configuration mechanisms,
either stateful or stateless, MUST NOT  enable multicast DNS.

































Esibov, Aboba & Thaler       Standards Track                   [Page 17]







to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 08:47:50 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA16983
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 08:47:49 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hrqd-000IRq-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:13:31 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hrqc-000IRk-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:13:30 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hrqc-0003WO-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:13:30 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Robert Elz <kre@munnari.OZ.AU>
To: Bernard Aboba <aboba@internaut.com>
cc: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language 
In-Reply-To: <Pine.BSF.4.21.0109130920090.1339-100000@internaut.com> 
References: <Pine.BSF.4.21.0109130920090.1339-100000@internaut.com> 
Date: Fri, 14 Sep 2001 09:28:27 +0700
Message-ID: <3399.1000434507@brandenburg.cs.mu.OZ.AU>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

    Date:        Thu, 13 Sep 2001 09:23:21 -0700 (PDT)
    From:        Bernard Aboba <aboba@internaut.com>
    Message-ID:  <Pine.BSF.4.21.0109130920090.1339-100000@internaut.com>

  | I would like to propose the following text on mDNS configuration be added
  | to the next version of the draft, and that previous text relating to use
  | of the domain searchlist for configuration be removed. Comments welcome. 

In general, this looks fine to me.   But there is one issue that isn't
clear (I suspect the problem is just wording) ...

  | Multicast DNS usage can be configured manually or automatically.  Where
  | no manual or automatic configuration is provided, multicast DNS is
  | enabled by default.

If multicast DNS exists, then there ought be (regardless of what other
mechanisms exist) a manual way of configuring it.   So I can't see a case
where "no manual configuration is provided" - maybe you mean "where no
manual or automatic configuration has been performed" ?   That is, what
matters is whether the switch has been set, rather than whether the switch
exists?

  | If a host is configured via automatic configuration mechanisms, either
  | stateful or stateless, and multicast DNS is not explicitly enabled, then
  | multicast DNS MUST NOT be used,

Does "is configured" there mean "multicast DNS has been configured" or
the host has been configured in any way at all?   Some people consider
deciding how many discs to buy as "configuring the host" ...   Does IPv6
auto-configuration of a link local address (which every IPv6 node will do)
count as "configured by automatic means" ?   An IPv6 host on a net with
no other nodes at all can get that far - so can two nodes on an isolated
net.   I suspect the wording needs a little tightening.

kre



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 08:47:57 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA16995
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 08:47:57 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hrkA-000IFa-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:06:50 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hrk9-000IFU-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:06:49 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hrk8-0003Vo-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:06:48 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BA1575B.A35705FF@email.mot.com>
Date: Fri, 14 Sep 2001 11:03:23 +1000
From: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
To: Bernard Aboba <aboba@internaut.com>
CC: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
References: <Pine.BSF.4.21.0109130920090.1339-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Bernard Aboba wrote:
> 
> I would like to propose the following text on mDNS configuration be added
> to the next version of the draft, and that previous text relating to use
> of the domain searchlist for configuration be removed. Comments welcome.
> 

As I have said before, I am in favour of removing the domain
searchlist approach for mDNS configuration.

> 3.1.  mDNS configuration
> 
> Multicast DNS usage can be configured manually or automatically.  Where
> no manual or automatic configuration is provided, multicast DNS is
> enabled by default.
> 

"Configured manually or automatically" could be tighter.
Don't we really mean that mDNS shouldn't be turned on if there is
a "configuration protocol" in operation (e.g. DHCP, stateless autoconfig)

"MAY be enabled by default", or "MUST be enabled by default".
I would prefer MAY be enabled by default.

Suggested text:
  Multicast DNS operation may be configured via DHCP, stateless DNS
  discovery or some other configuration protocol.  Where no configuration
  protocol is in operation, multicast DNS MAY be enabled by default.

> For IPv6, the stateless DNS discovery mechanisms described in [19] can
> be used to discover whether multicast DNS is enabled or disabled.
> 
> Where DHCPv4 or DHCPv6 is implemented, DHCP options can be used to
> configure multicast DNS. The mDNS enable DHCP option, described in [6],
> can be used to explicitly enable or disable use of multicast DNS. The
> Name Service Search option, described in RFC 2937 [3], can be used to
> determine where multicast DNS is used within the name service search
> order. DHCP option codes are used as RFC 2937 codes signifying name
> services within the search order. As a result, to specify multicast DNS
> usage within the name service search order, the option code assigned to
> the mDNS enable option is used.
> 

As it happens, I just submitted an I-D intended to update rfc2937.
It describes an alternate way of doing this, without a new DHCP
option.  This draft should make its way through the system, but
for now I've put a copy up on the following URL:

  http://internet.motlabs.com/~aidan/draft-williams-dhc-rfc2937bis-00.txt

I was under the (possibly wrong) assumption that DHCP options were
in reasonably short supply, and because mDNS doesn't require IP
address configuration we probably didn't need one.

The relevant text to mDNS enablement in draft-...-rfc2937bis:

   6.1.  Enabling/Disabling mDNS

      The presence of the mDNS name service search option in the search
      list enables the use of multicast DNS on a host.  A host using
      DHCP SHOULD NOT enable multicast DNS unless the mDNS name service
      is listed in the Name Service Search Option list.

      An alternative approach is to have a seperate DHCP option to
      enable/disable mDNS.  If mDNS is to be used with multicast scopes
      larger than link-local, it is reasonable to be able to specify the
      order in which different scopes are consulted.  For example using
      a search path of mdns-link-local, netbios, mdns-site-local, isp-
      dns.

This text probably belongs somewhere other than an rfc2937bis draft,
and is slightly out of order..

rfc2937bis advantages/disadvantages
  - support additional mdns multicast scopes cleanly
  - presence of mDNS scope number in the DHCP name service search
    list option disables/enables mDNS for that interface, using
    that multicast scope
  - enabling mdns in a DHCP-configured environment is tied to
    implementation of name service search option in the host,
    which is probably larger/harder to implement than the
    mdns-enable approach.

new DHCP option (mdns-enable) + search path
  - supporting multiple scopes isn't clean.  I think it is
    reasonable to support multiple scopes, but others may disagree.
  - consumes a dhcp option, and as currently specified makes an
    additional search-list-like option
  - easier to implement..
  - need to update rfc2937 anyway because of text like this:
      "we rely on the existence of previously-defined DHCP options
       that specify the IP address(es) of servers which provide
       name services whose order we wish to express."
    mdns-enable doesn't specify an IP address (currently at least).

> If a host is configured via automatic configuration mechanisms, either
> stateful or stateless, and multicast DNS is not explicitly enabled, then
> multicast DNS MUST NOT be used, ensuring that upgraded hosts do not
> change their default behavior.  This implies that the host will neither
> listen on the DNS LINKLOCAL multicast address, nor will it send queries
> to that address.  For a DNS server, automatic configuration mechanisms,
> either stateful or stateless, MUST NOT  enable multicast DNS.
> 

Is this a global disable on the host, or a per-interface enable/disable?
My opinion is that it should be per-interface.

regards
	aidan
____
:wq!

Communications and Networks Lab         aidan.williams@motorola.com
Motorola Australian Research Centre          phone: +61 2 9666 0649
Locked Bag 5028, Botany NSW 1455              phax: +61 2 9666 0501
Australia                            http://cnlab.arc.corp.mot.com/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 08:53:07 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA17069
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 08:53:06 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hs3t-000Iqc-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:27:13 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hs3t-000IqW-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:27:13 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hs3s-0003Wk-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:27:12 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 13 Sep 2001 20:22:19 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Robert Elz <kre@munnari.OZ.AU>
cc: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language 
In-Reply-To: <3399.1000434507@brandenburg.cs.mu.OZ.AU>
Message-ID: <Pine.BSF.4.21.0109132015290.2003-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>   | Multicast DNS usage can be configured manually or automatically.  Where
>   | no manual or automatic configuration is provided, multicast DNS is
>   | enabled by default.
> 
> If multicast DNS exists, then there ought be (regardless of what other
> mechanisms exist) a manual way of configuring it.   So I can't see a case
> where "no manual configuration is provided" - maybe you mean "where no
> manual or automatic configuration has been performed" ?   That is, what
> matters is whether the switch has been set, rather than whether the switch
> exists?
> 

Yes, "where no manual or automatic configuration has been performed" was
the intent. So if you don't have a DHCP server or stateless autoconfig 
then mDNS will be enabled. 

>   | If a host is configured via automatic configuration mechanisms, either
>   | stateful or stateless, and multicast DNS is not explicitly enabled, then
>   | multicast DNS MUST NOT be used,
> 
> Does "is configured" there mean "multicast DNS has been configured" or
> the host has been configured in any way at all?   

If the host is configured via DHCP or stateless autoconfig, and the
administrator does not configure mDNS, the assumption is that it will not
be used. The goal is for hosts to exhibit their present behavior (no
mDNS) within configured environments, unless the administrator explicitly
enables it. 

> Does IPv6 auto-configuration of a link local address (which every IPv6
> node will do) count as "configured by automatic means" ?  

No, but implementation of router advertisement and DNS discovery would. 

> I suspect the wording needs a little tightening.

Suggestions welcome. 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 09:00:10 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA17185
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 09:00:09 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hs7z-000IyA-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:31:27 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hs7y-000Iy2-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:31:26 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hs7y-0003Wy-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:31:26 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BA19266.48A83FCC@email.mot.com>
Date: Fri, 14 Sep 2001 15:15:18 +1000
From: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
To: Bernard Aboba <aboba@internaut.com>
CC: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
References: <Pine.BSF.4.21.0109132022300.2003-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Bernard Aboba wrote:
> 
> > As it happens, I just submitted an I-D intended to update rfc2937.
> > It describes an alternate way of doing this, without a new DHCP
> > option.  This draft should make its way through the system, but
> > for now I've put a copy up on the following URL:
> >
> >   http://internet.motlabs.com/~aidan/draft-williams-dhc-rfc2937bis-00.txt
> 
> RFC 2937 only talks about configuring the order in which naming protocols
> are used. The document states that interpretation of the option is
> "implementation dependent", so that it is not clear that the name search
> list is to be used for enabling/disabling naming protocols.
> 

I agree that the current spec is not compelling on this point.
I'm suggesting a change to the RFC 2937 behaviour, possibly
specifically for mDNS.

A host can ignore part or all of a name service search list, but the
absence or presence of a particular name service type in a supplied
search list surely suggests whether it should be used or not.

> > rfc2937bis advantages/disadvantages
> >   - support additional mdns multicast scopes cleanly
> 
> mDNS enable/disable option can support multiple scopes, no?
> 

Yes, but not in this way:
  mdns-link-local, netbios, mdns-site-local, isp-dns

I view mdns-enable as an additional search-list proposal, to a large
extent overlapping rfc2937, and without the ability to interleave
different mdns multicast scopes with other name services (as above).

Maybe the real problem is that nobody wants to (or has) implemented
the host support for rfc2937 as is.

> >   - presence of mDNS scope number in the DHCP name service search
> >     list option disables/enables mDNS for that interface, using
> >     that multicast scope
> 
> Careful. RFC 2937 does not state that the name service search list is to
> be interpreted this way.
> 

See above.

> >   - enabling mdns in a DHCP-configured environment is tied to
> >     implementation of name service search option in the host,
> >     which is probably larger/harder to implement than the
> >     mdns-enable approach.
> 
> Yes, the changes required to implement this option are considerably
> greater than to support the mDNS enable/disable option.
> 

Assuming that the rfc2937 searchlist was not already implemented.

> >   - consumes a dhcp option, and as currently specified makes an
> >     additional search-list-like option
> 
> RFC 2937 appears to require assignment of a DHCP option code as a
> prerequisite.
> 

Yes, and that is why I'm suggesting an update to it.  The draft I
wrote does not require a new DHCP option code to be defined, at the
expense of a new IANA number space for the searchlist.

regards
	aidan
____
:wq!

Communications and Networks Lab         aidan.williams@motorola.com
Motorola Australian Research Centre          phone: +61 2 9666 0649
Locked Bag 5028, Botany NSW 1455              phax: +61 2 9666 0501
Australia                            http://cnlab.arc.corp.mot.com/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 09:00:15 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA17163
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 08:59:50 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hs8H-000Iyc-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:31:45 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hs8G-000IyW-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:31:44 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hs8G-0003X0-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:31:44 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 13 Sep 2001 22:18:01 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
cc: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
In-Reply-To: <3BA19266.48A83FCC@email.mot.com>
Message-ID: <Pine.BSF.4.21.0109132215170.2136-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> > mDNS enable/disable option can support multiple scopes, no?
> > 
> 
> Yes, but not in this way:
>   mdns-link-local, netbios, mdns-site-local, isp-dns

Yes, RFC 2937 can provide the *order* in which the services are used.

> I view mdns-enable as an additional search-list proposal, to a large
> extent overlapping rfc2937, and without the ability to interleave
> different mdns multicast scopes with other name services (as above).
> 

Not sure I agree with that. mdns-enable doesn't overlap with RFC 2937,
since that document isn't clear about enable/disable semantics. 

> Maybe the real problem is that nobody wants to (or has) implemented
> the host support for rfc2937 as is.

Is that true? No implementations at all? 




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 09:01:41 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA17221
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 09:01:40 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hs4D-000Iqr-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:27:33 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hs4D-000Iql-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:27:33 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hs4C-0003Wn-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:27:32 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 13 Sep 2001 20:32:45 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
cc: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
In-Reply-To: <3BA1575B.A35705FF@email.mot.com>
Message-ID: <Pine.BSF.4.21.0109132022300.2003-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> As it happens, I just submitted an I-D intended to update rfc2937.
> It describes an alternate way of doing this, without a new DHCP
> option.  This draft should make its way through the system, but
> for now I've put a copy up on the following URL:
> 
>   http://internet.motlabs.com/~aidan/draft-williams-dhc-rfc2937bis-00.txt

RFC 2937 only talks about configuring the order in which naming protocols
are used. The document states that interpretation of the option is
"implementation dependent", so that it is not clear that the name search
list is to be used for enabling/disabling naming protocols. 

> rfc2937bis advantages/disadvantages
>   - support additional mdns multicast scopes cleanly

mDNS enable/disable option can support multiple scopes, no?

>   - presence of mDNS scope number in the DHCP name service search
>     list option disables/enables mDNS for that interface, using
>     that multicast scope

Careful. RFC 2937 does not state that the name service search list is to
be interpreted this way. 

>   - enabling mdns in a DHCP-configured environment is tied to
>     implementation of name service search option in the host,
>     which is probably larger/harder to implement than the
>     mdns-enable approach.

Yes, the changes required to implement this option are considerably
greater than to support the mDNS enable/disable option. 

> new DHCP option (mdns-enable) + search path
>   - supporting multiple scopes isn't clean.  I think it is
>     reasonable to support multiple scopes, but others may disagree.

mDNS enable option can support multiple scopes, too.

>   - consumes a dhcp option, and as currently specified makes an
>     additional search-list-like option

RFC 2937 appears to require assignment of a DHCP option code as a
prerequisite.

>   - easier to implement..

Yes. 

>   - need to update rfc2937 anyway because of text like this:
>       "we rely on the existence of previously-defined DHCP options
>        that specify the IP address(es) of servers which provide
>        name services whose order we wish to express."

>     mdns-enable doesn't specify an IP address (currently at least).
> 
> Is this a global disable on the host, or a per-interface enable/disable?
> My opinion is that it should be per-interface.
> 

I think that is correct. 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 09:01:51 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA17237
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 09:01:50 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hs9P-000J15-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:32:55 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hs9O-000J0z-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:32:54 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hs9N-0003X8-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:32:53 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BA1A085.5C22541C@email.mot.com>
Date: Fri, 14 Sep 2001 16:15:33 +1000
From: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
To: Bernard Aboba <aboba@internaut.com>
CC: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
References: <Pine.BSF.4.21.0109132215170.2136-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Bernard Aboba wrote:
> 
> > > mDNS enable/disable option can support multiple scopes, no?
> > >
> >
> > Yes, but not in this way:
> >   mdns-link-local, netbios, mdns-site-local, isp-dns
> 
> Yes, RFC 2937 can provide the *order* in which the services are used.
> 

The current mdns enable draft has this diagram:

  2.0 DHCP Enable mDNS Option Format

       Code   Len  Scope
      +-----+-----+-----+-----+
      | TBD |  n  | Scp | ...
      +-----+-----+-----+-----+

The DHCP option number is TBD.  Support is there for additional
multicast scopes should they be specified.

RFC 2937 would then use the option number TBD in the search list.

But this option number identifies *all* of the mDNS scopes which
may be consulted (presumably they are queried one after another).
Thus different mDNS scopes cannot be interleaved with other name
services, as in the example I gave above.

I think this is useful functionality, others may disagree.


> > I view mdns-enable as an additional search-list proposal, to a large
> > extent overlapping rfc2937, and without the ability to interleave
> > different mdns multicast scopes with other name services (as above).
> >
> 
> Not sure I agree with that. mdns-enable doesn't overlap with RFC 2937,
> since that document isn't clear about enable/disable semantics.
> 

I would argue that the lack of clarity in rfc2937 about enable/disable
is a problem with the specification.  Enable/disable seems to be
implied by specifying a list of name services, but the text avoids
the issue.


> > Maybe the real problem is that nobody wants to (or has) implemented
> > the host support for rfc2937 as is.
> 
> Is that true? No implementations at all?

I haven't come across it in real life, but I'm not sure.  I was posing
the question to get some idea for the following reason:

If rfc2937 is little supported, then using it to enable mDNS is going
to be a large hill to climb -- even if it seems like a sensible
approach.

regards
	aidan
____
:wq!

Communications and Networks Lab         aidan.williams@motorola.com
Motorola Australian Research Centre          phone: +61 2 9666 0649
Locked Bag 5028, Botany NSW 1455              phax: +61 2 9666 0501
Australia                            http://cnlab.arc.corp.mot.com/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 09:08:51 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA17305
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 09:08:51 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hsGo-000JII-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:40:34 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hsGn-000JHw-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:40:33 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hsGm-0003Xp-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:40:32 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Robert Elz <kre@munnari.OZ.AU>
To: Bernard Aboba <aboba@internaut.com>
cc: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language 
In-Reply-To: <Pine.BSF.4.21.0109132015290.2003-100000@internaut.com> 
References: <Pine.BSF.4.21.0109132015290.2003-100000@internaut.com> 
Date: Fri, 14 Sep 2001 14:03:34 +0700
Message-ID: <1279.1000451014@brandenburg.cs.mu.OZ.AU>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

    Date:        Thu, 13 Sep 2001 20:22:19 -0700 (PDT)
    From:        Bernard Aboba <aboba@internaut.com>
    Message-ID:  <Pine.BSF.4.21.0109132015290.2003-100000@internaut.com>

  | Yes, "where no manual or automatic configuration has been performed" was
  | the intent. So if you don't have a DHCP server or stateless autoconfig 
  | then mDNS will be enabled.

Yes, that's what I thought, so change that wording (to "has been performed").

  | If the host is configured via DHCP or stateless autoconfig, and the
  | administrator does not configure mDNS, the assumption is that it will not
  | be used.

Yes, that's what I thought - the point was that "configured" is a very
broad term, and can apply to almost anything.

  | No, but implementation of router advertisement and DNS discovery would. 

Not implementation, surely, only actually receiving a router advertisement.

  | > I suspect the wording needs a little tightening.
  | Suggestions welcome. 

For this one, perhaps

	If a host has been configured via any automatic configuration
	mechanism which is able to supply DNS configuration information,
	then multicast DNS MUST NOT be used unless it has been explicitly
	enabled, whether via that mechanism or any other.  This ensures that
	upgraded hosts do not change their default behaviour, without
	requiring the source of the configuration information to be
	simultaneously updated.  This implies....

kre



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 14 09:13:20 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA17426
	for <dnsext-archive@lists.ietf.org>; Fri, 14 Sep 2001 09:13:19 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15hsJ2-000JNX-00
	for namedroppers-data@psg.com; Fri, 14 Sep 2001 05:42:52 -0700
Received: from h-135-207-10-108.research.att.com ([135.207.10.108] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15hsJ2-000JNP-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 05:42:52 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15hsJ1-0003YC-00
	for namedroppers@ops.ietf.org; Fri, 14 Sep 2001 08:42:51 -0400
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BA1AD76.3A6C60E@email.mot.com>
Date: Fri, 14 Sep 2001 17:10:46 +1000
From: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
To: Bernard Aboba <aboba@internaut.com>
CC: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
References: <Pine.BSF.4.21.0109132215170.2136-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Bernard Aboba wrote:
> > Maybe the real problem is that nobody wants to (or has) implemented
> > the host support for rfc2937 as is.
> 
> Is that true? No implementations at all?

Most freebie unixes would use one of these two:
  - The ISC dhcp client appears not to, looking at the source.
  - pump appears not to, looking at the source.

Ironically, the Solaris startup files I have examined edit
/etc/nsswitch.conf on the basis of whether there was a DNS
server IP address returned in the DHCP response, so I
conclude it is not supported.

Not sure about windows, but I'm sure someone knows.. ;)

regards
	aidan
____
:wq!

Communications and Networks Lab         aidan.williams@motorola.com
Motorola Australian Research Centre          phone: +61 2 9666 0649
Locked Bag 5028, Botany NSW 1455              phax: +61 2 9666 0501
Australia                            http://cnlab.arc.corp.mot.com/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 01:15:13 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA29420
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 01:15:13 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15iTd0-000OND-00
	for namedroppers-data@psg.com; Sat, 15 Sep 2001 21:33:58 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15iTcz-000ON3-00
	for namedroppers@ops.ietf.org; Sat, 15 Sep 2001 21:33:57 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15iTcz-000JJ0-00
	for namedroppers@ops.ietf.org; Sat, 15 Sep 2001 21:33:57 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109152145.OAA02355@scv3.apple.com>
Subject: Re: Sense of the WG: mDNS configuration?
Date: Sat, 15 Sep 2001 14:45:15 -0700
From: Stuart Cheshire <cheshire@apple.com>
To: <namedroppers@psg.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Just when I think we've all agreed that it's a bad idea to overload the 
DNS search list to configure mDNS, someone always seems to suggest it 
again.

In case there is anyone still in favour of that alternative, let me make 
the following comment:

>Currently the DNS server addresses and the domain searchlist can be
>statelessly discovered in IPv6, but the other candidates cannot be,
>thereby creating a DHCPv6 dependency, which is a show-stopper.

I really believe this is misleading reasoning.

Yes, it is true that discovering the domain search list statelessly in 
IPv6 avoids the need for a DHCP server, but if we overload the DNS search 
list to control this aspect of host configuration, where do we stop? Do 
we configure the default printer by adding an entry to the search list 
with the name "your.default.printer.is.printer1.ietf.org."? I could 
easily define an encoding to let you embed *any* legal DHCP option as an 
entry in the search list.

Maybe it would be a good idea to just get rid of DHCPv6 and use some 
DNS-based solution to retrieve host configuration information, but that 
would be a much larger discussion, not some ad-hoc hack just for this 
specific purpose.

Stuart Cheshire <cheshire@apple.com>
 * Wizard Without Portfolio, Apple Computer
 * Chairman, IETF ZEROCONF
 * www.stuartcheshire.org




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 08:52:37 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA11461
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 08:52:37 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ib2y-0003bp-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 05:29:16 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ib2x-0003bK-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:29:15 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ib2x-0005iX-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:29:15 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109160825.BAA01020@scv1.apple.com>
Subject: Re: Sense of the WG: mDNS configuration?
Date: Sun, 16 Sep 2001 01:25:18 -0700
From: Stuart Cheshire <cheshire@apple.com>
To: "Bernard Aboba" <aboba@internaut.com>
cc: <namedroppers@psg.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>I'll let someone else comment on whether the proposals are an "ad-hoc
>hack", but the DNS Discovery work is not motivated by the desire to
>configure mDNS. Rather the goal is to configure statelessly configure DNS. 

You misunderstood me. I agree that DNS Discovery is useful.

My comment about "ad-hoc hack" referred solely to using magic strings in 
the search list as a back-door way to get configuration information to 
clients.

Stuart Cheshire <cheshire@apple.com>
 * Wizard Without Portfolio, Apple Computer
 * Chairman, IETF ZEROCONF
 * www.stuartcheshire.org




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 08:52:38 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA11462
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 08:52:37 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15iaz9-0003Yf-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 05:25:19 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15iaz8-0003YZ-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:25:18 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15iaz8-0005bn-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:25:18 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109160627.XAA05924@scv3.apple.com>
Subject: Re: New mDNS configuration language
Date: Sat, 15 Sep 2001 23:27:12 -0700
From: Stuart Cheshire <cheshire@apple.com>
To: <namedroppers@ops.ietf.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

I think that before we can have a meaningful discussion about how to 
configure mDNS, we need to agree on what we *mean* by "mDNS".

There are currently two different things being called "mDNS", broadly 
similar, but different in one small though important way:

1. Use of IP multicast to look up DNS RRs that have names in the global 
DNS name space, when there is no access to the global tree of 
authoritative DNS servers which manage that global name space via formal 
delegation.

2. Use of IP multicast to look up DNS RRs that have names within a 
portion of the DNS name space specially designated to have only local 
significance, akin to Net 10 IPv4 addresses, or 169.254/16 link-local 
IPv4 addresses.

Version 1 is useful in the situation when my office uplink to the rest of 
the network has failed, and I double-click an alias icon on my laptop 
computer to access the disk of my desktop machine sitting next to it. The 
alias icon contains the DNS name of the target host (chesh1.apple.com.) 
so today this would simply fail because it couldn't get to the DNS 
server. If we had multicast DNS, then perhaps it could resolve the name 
locally using multicast, and succeed. There is also a risk here that 
without DNS SEC, the client may not know whether the response is 
trustworthy.

Version 2 is useful for a home user who simply wants to name the devices 
on his or her home network, without the expense or hassle of having to 
pay some registry to "buy" some chunk of the global DNS name space within 
which he or she can allocate names. This is much like what users do today 
with AppleTalk NBP names, or Microsoft NetBIOS names.

For Version 1, an option to enable and disable it seems to make sense.
It might also make sense to have an option to conditionally enable it, 
only for responses authenticated via DNS SEC, or universally enable it, 
for all responses whether authenticated or not.

For Version 2, an option to enable and disable makes much less sense.
If the designated local name space is "local.arpa." and a user types 
"myprinter.local.arpa." into their printing client, and there is an 
mDNS-capable printer on the network willing to answer to that name, then 
what possible utility is gained by having the host software decide to 
deliberately fail? There's no DHCP option to "disable" Net 10 addresses 
-- if you don't want Net 10 addresses, then you simply don't use them. If 
you do decide to use Net 10 addresses, the last thing you need is to 
struggle with them not working because some other configuration option 
somewhere else in the system has disabled them. The same applies here. No 
one has to use names within the designated local name space if they don't 
want to. If someone *does* decide to use names within the designated 
local name space, let's not work to create strange and interesting ways 
for them to not work, to confuse and bewilder our poor end-users.

Stuart Cheshire <cheshire@apple.com>
 * Wizard Without Portfolio, Apple Computer
 * Chairman, IETF ZEROCONF
 * www.stuartcheshire.org




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 08:52:40 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA11482
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 08:52:40 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ib19-0003aI-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 05:27:23 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ib19-0003aC-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:27:23 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ib19-0005fL-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:27:23 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Sun, 16 Sep 2001 00:26:59 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Stuart Cheshire <cheshire@apple.com>
cc: namedroppers@psg.com
Subject: Re: Sense of the WG: mDNS configuration?
In-Reply-To: <200109152145.OAA02355@scv3.apple.com>
Message-ID: <Pine.BSF.4.21.0109160020200.7247-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Just when I think we've all agreed that it's a bad idea to overload the 
> DNS search list to configure mDNS, someone always seems to suggest it 
> again.

Actually, we haven't heard all that many arguments in favor of the domain
searchlist as of late. So you may be preaching to the choir ;)

> Yes, it is true that discovering the domain search list statelessly in 
> IPv6 avoids the need for a DHCP server, but if we overload the DNS search 
> list to control this aspect of host configuration, where do we stop? 

Actually, I don't believe that is the approach that DNS Discovery is
taking, although there *is* a proposal for a general method of encoding
parameters in a (new? TXT?) DNS RR. 

> Maybe it would be a good idea to just get rid of DHCPv6 and use some 
> DNS-based solution to retrieve host configuration information, but that 
> would be a much larger discussion, not some ad-hoc hack just for this 
> specific purpose.

I'll let someone else comment on whether the proposals are an "ad-hoc
hack", but the DNS Discovery work is not motivated by the desire to
configure mDNS. Rather the goal is to configure statelessly configure DNS. 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 08:52:44 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA11494
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 08:52:43 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ib2n-0003bI-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 05:29:05 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ib2n-0003bC-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:29:05 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ib2n-0005iF-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 05:29:05 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109160822.BAA11508@scv3.apple.com>
Subject: Re: Question on Stewart's last mDNS item
Date: Sun, 16 Sep 2001 01:22:06 -0700
From: Stuart Cheshire <cheshire@apple.com>
To: "Matt Crawford" <crawdad@fnal.gov>, <namedroppers@ops.ietf.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>Did svrloc crash & burn without me noticing?  Or is there a
>terrible gap in coverage?

A while back I promised you an answer to this question. It took me a 
while, but here it is.

I realized that one goal of Multicast DNS was to replace AppleTalk NBP, 
but no one had actually written down precisely what is required to 
replace AppleTalk NBP. To fill that gap, I have written a draft, which I 
just submitted to Internet-Drafts as draft-cheshire-dnsext-nbp-00.txt.

This file can also be fetched via http from:
<http://files.multicastdns.org/draft-cheshire-dnsext-nbp.txt>

The quick summary is that the main point of NBP is not service discovery 
or network browsing. The main point of NBP is that it defines a name 
space which is free for local shared use, without coordination with any 
registry. You are free to use any name you please on the local network. 
The only caveat is that so is anyone else, so you'd better play nicely 
together.

I enclose the core of the text here for convenience.

11 requirements for an IP-based replacement for AppleTalk NBP:

1 Name-to-Address Mapping

   NBP's primary function is translating names to addresses.

   NBP stands for Name Binding Protocol, not Network Browsing Protocol.
   Many people know NBP only as "that thing that lets you browse the
   network in the Macintosh Chooser". While browsing is an important
   facility of NBP, it is secondary to NBP's primary function of
   translating names to addresses.

   Every time a user prints using AppleTalk, the printing software takes
   the name of the currently selected printer, looks up the current
   AppleTalk address associated with that named service, and establishes
   a connection to that service on the network. The user may invoke
   NBP's browsing capability once when first selecting the desired
   printer in the Chooser, but then after that, every single time they
   print anything, it is a simple efficient name-to-address lookup that
   is being performed, not a full-fledged browsing operation.

   Any NBP replacement needs to support, as it's primary function,
   an efficient name-to-address lookup operation.

2 Name Services, not Hardware

   The primary named entities in NBP are services, not "hosts",
   "machines", "devices", or pieces of hardware of any kind. This
   concept is more subtle than it may seem at first, so it bears some
   discussion.

   The AppleTalk NBP philosophy is that naming a piece of hardware on
   the network is of little use if you can't communicate with that piece
   of hardware. To communicate with a piece of hardware, there needs to
   be a piece of software running on that hardware which sends and
   receives network packets conforming to some specific protocol. This
   means that whenever you communicate with a machine, you are really
   communicating with some piece of software on that machine. Even if
   you just 'ping' a machine to see if it is responding, it is not
   really the machine that you are 'pinging', it is the software on that
   machine that generates ICMP Echo Responses.

   Consequently, this means that the only thing worth naming is the
   software entities with which you can communicate. A user who wants to
   use a print server or a file server needn't care about what hardware
   implements those services. There may be a single machine hosting both
   services, or there may be two separate machines. The end user doesn't
   need to care.

   The one exception to this is network managers, who may want to name
   physical hardware for the purpose of tracking physical inventory.
   However, even this can be recast into a service-oriented view of the
   world by saying that what you're naming is not the hardware, but the
   ICMP Echo Responder that runs (or is assumed to be running) on every
   piece of IP hardware.

3 Address Services, not Hardware
    -or-
    Escape the Tyranny of Well Known Ports

   The reader may argue that DNS already supports the philosophy of
   naming services instead of hosts. When we see names like
   "www.example.com.", "pop.example.com.", "smtp.example.com.",
   "news.example.com." and "time.example.com.", we do not assume that
   each of those names refer to a different host. They are clearly
   intended to be logical service names, and could in fact all refer to
   the same IP address.

   The shortcoming here is that although the names are clearly logical
   service names, the result today of doing a conventional ("A" Record)
   DNS lookup for those names gives you only the IP address of the
   hardware where the service is located. To communicate with the
   desired service, you also need to know the TCP or UDP port number at
   which the service can be reached, not just the IP address.

   This means that the port number has to be communicated out-of-band,
   in some other way. One way is for the port number to be a specific
   well-known constant for any given protocol. This makes it hard to
   run more than one instance of a service on a single piece of
   hardware. Another way is for the user to explicitly type in the port
   number, for example, "www.example.com.:8080" instead of
   "www.example.com.", but needing to know and type in a port number is
   as ugly and fragile as needing to know and type in an IP address.

   Another aspect of the difficulty of running more than one instance of
   a service on a single piece of hardware is that it forces application
   programmers to write their own demultiplexing capability. For
   example, if an AppleTalk print server offers three print queues, each
   print queue runs as its own independent service, listening on its own
   port number (called a socket number in AppleTalk terminology) and
   each is advertised as a separate independent named NBP entity. When a
   client looks up the address of that named NBP entity, the reply
   encodes not only on which net and subnet the service resides, and on
   which host on that subnet (like an IP address does), but also on
   which port number (socket number) within that host. In contrast, if
   an lpr print server offers three print queues, all three print queues
   are typically reached through the same well-known port number, and
   then the lpr protocol has to use it own demultiplexing capability
   (the print queue name) in order to determine which print queue is
   sought. This makes it especially difficult to run two different
   pieces of print queue software from different vendors on the same
   machine, because they cannot both use the same well-known port.

   A similar trick is used in HTTP 1.1, where the Host header is used to
   allow multiple logical http services to run at the same IP address.
   Again, this works for a single-vendor solution, but if you have an
   image server, a database program, an http email access gateway, and a
   regular http server, they can't all run on the same TCP port on the
   same machine.

   Yet another problem of well-known ports is that port numbers are a
   finite resource. Originally, port numbers 0-1023 were reserved for
   well-known services, and the remaining 98% of the port space was free
   for dynamic allocation. Since then, the range of "Registered Ports"
   has crept upwards until today, ports 0-49151 are reserved, and only
   25% of the space remains available for dynamic allocation. Even
   though 65535 may seem like a lot of available port numbers, with the
   pace of software development today, if every new protocol gets its
   own private port number, we will eventually run out. To avoid having
   to do application-level demultiplexing, protocols like the X Window
   System wisely use a range of port numbers, and let TCP do the
   demultiplexing for them. The X Window System uses 64 ports, in the
   range 6000-6063. If every new protocol were to get its own chunk of
   64 ports, we would run out even faster.

   Any NBP replacement needs to provide, not just the network number,
   subnet number, and host number within that subnet (i.e. the IP
   address) but also the port number within that host where the service
   is located. Furthermore, since many existing IP services such as lpr
   *do* already use additional application-layer demultiplexing
   information such as a print queue name, an NBP replacement needs to
   support this too by including this information as part of the
   complete package of addressing information provided to the client to
   enable it to use the service. The NBP replacement needs to name
   individual print queues as first-class entities in their own right.
   It is not sufficient to name a print server, within which separate
   print queues can be found.

   One possible answer here is that an IP-based NBP replacement could
   use a solution derived from DNS "SRV" records instead of "A" records,
   since SRV records *do* provide a port number. However, this alone is
   not a complete solution, because "SRV" records cannot tell you an lpr
   print queue name.

4 Typed Name Space

   AppleTalk NBP names are structured names, of the form:

      Name : Type @ Zone

   The Name is the user-visible name of the service.

   The Type is an opaque identifier which identifies the type of
   service. For convenience, the opaque identifier is generally
   constructed using descriptive ASCII text, but this text has no
   meaning to the protocol, and care should be taken in inferring too
   much meaning from it. For example, the NBP Service Type "LaserWriter"
   means "any service that speaks PostScript over AppleTalk Printer
   Access Protocol". It does not necessarily mean an Apple-branded
   "LaserWriter" printer; nor does the service even have to be a
   printer. A device that archives documents to recordable CDs could
   advertise itself as a "LaserWriter", meaning that it speaks
   PostScript over PAP, not necessarily that it prints that PostScript
   on paper when it gets it. The end-user never directly sees the
   Service Type. It is implicit in the user's action; e.g. when
   printing, the printing software knows what protocol(s) it speaks and
   consequently what Service Type(s) it should be looking for -- the
   user doesn't have to tell it.

   The Zone is an organizational or geographical grouping of named
   services. Typical AppleTalk Zone Names are things like "Engineering"
   and "Sales". The equivalent concept in DNS could be a subdomain such
   as "engineering.company.com." or "sales.company.com."

   Each {Type,Zone} pair defines a name space in which service names can
   be registered. It is not a name conflict to have a printer called
   "Sales" and a file server called "Sales", because one is
   "Sales:LaserWriter@Zone" and the other is "Sales:AFPServer@Zone".

   Any NBP replacement needs to provide a mechanism that allows names to
   be grouped into organizational or geographical "zones", and within
   each "zone", to provide an independent name space for each service
   type.

5 User-Friendly Names

   When repeatedly typing in names on command-line systems, it is
   helpful to have names that are short, all lower-case, with no spaces
   or other unusual characters.

   Since Service Names are intended to be selected from a list, not
   repeatedly typed in on a keyboard, there is no for them to be
   restricted so. Users should be able to give their printers names like
   "Sales", "Marketing", and "3rd Floor Copy Room", not just
   "printer1.ietf.org." Of course a user is free to restrict their
   Service Names to lower-case letters without spaces if they wish, but
   they should not be forced to do that.

   Any NBP replacement needs to support a full range of rich text
   characters, including upper case, lower case, spaces, accented
   characters, and so on. The correct solution is likely to be Unicode,
   probably in the form of UTF-8.

   Note that although the characters ':' and '@' are used when writing
   AppleTalk NBP names, they are simply a notational convenience in
   written text. In the on-the wire protocol and in the software data
   structures, NBP Name, Type and Zone strings are all allowed to
   contain arbitrary characters, including ':' and '@'.

6 Zeroconf Operation

   AppleTalk NBP is self-configuring. On a network of just two hosts,
   they communicate peer-to-peer using multicast. On a large managed
   network, AppleTalk routers automatically perform an aggregation
   function, allowing name lookups to be performed via unicast to a
   service running on the router, instead of by flooding the entire
   network with multicast packets to every host.

   Any NBP replacement needs to operate in the absence of external
   network infrastructure. It should also be able to take advantage of
   appropriate external network infrastructure, where present, to
   perform queries via unicast instead of multicast.

7 Name Space Management
    -or-
    Name Conflict Detection

   Because an NBP replacement needs to operate in a Zeroconf
   environment, it cannot be assumed that a central network
   administrator is managing the network. In a managed network normal
   administrative controls may apply, but in the Zeroconf case an NBP
   replacement must make it easy for users to name their devices as they
   wish, without the inconvenience or expense of having to seek
   permission or pay some organization like a domain name registry for
   the privilege. However, this ease of naming and freedom to choose any
   desired name means that two users may independently decide to run a
   personal file server on their laptop computers, and (unimaginitively)
   name it "My Computer". When these two users later attend the next
   IETF meeting and find themselves part of the same wireless network,
   there may be problems.

   Similarly, every Epson Ethernet printer may ship from the factory
   with its Service Name set to "Epson Printer". On a typical small home
   network where there is only one printer this is not a problem, but it
   could be a problem if two or more such printers are connected to the
   same network.

   Any NBP replacement needs to detect such conflicts, and handle them
   appropriately. In the case of the laptop computers, which have
   keyboards, screens, and human users, the software should display a
   message telling one or both users that they need to select a new
   name.

   In the case of the printers which have no keyboard or screen, the
   software should automatically select a new unique name, perhaps by
   appending an integer to the end of the existing name, e.g. "Epson
   Printer 2".

   Because of the potentially transient nature of connectivity on the
   wireless networks that are becoming more and more common, this name
   conflict detection needs to be an ongoing process. It is not
   sufficient to simply verify uniqueness of names for a few seconds
   during the boot process and then assume that the names will remain
   unique indefinitely.

8 Late Binding

   When the user selects their default printer, the software should not
   store the IP address and port number, but just the name. Then, every
   time the user prints, the software should look up the name to find
   the current IP address and port number for that service. This allows
   a named logical service to be moved from one piece of hardware to
   another without disrupting the user's ability to print to that named
   print service.

   On a network using DHCP or self-assigned link-local addresses, a
   device's IP address may change from day to day. By deferring binding
   of name to address until actual use, this allows the client to get
   the correct IP address at the time the service is used.

   Similarly, with a service using a dynamic port number instead of a
   fixed well-known port, the service may not get the same port number
   every time it is started or restarted. By deferring binding of name
   to port number until actual use, this allows the client to get the
   correct port number at the time the service is used.

9 Simplicity

   Any NBP replacement needs to be simple enough that vendors of even
   the cheapest ink-jet printer can afford to implement it in the
   device's limited firmware.

10 Network Browsing

   AppleTalk NBP offers certain limited wildcard functionality. For
   example, the service name "*" means "any name". This allows a client
   to perform an NBP lookup such as "*:LaserWriter@MyZone" and receive
   back in response a list of all the PAP (AppleTalk Printer Access
   Protocol) printers in the Zone called "MyZone".

   Any NBP replacement needs to allow a piece of software, such as a
   printing client, or a file server client, to enumerate all the named
   instances of services in a specified zone (domain) which speak its
   protocol(s).

11 Browsing and Registration Guidance

   AppleTalk NBP provides certain meta-information to the client.

   On a network with multiple AppleTalk Zones, the AppleTalk network
   infrastructure informs the client of the list of Zones that are
   available for browsing. It also informs the client of the default
   Zone, which defines the client's logical "home" location. This is the
   Zone that is selected by default when the Macintosh Chooser is
   opened, and is usually the Zone where the user is most likely to find
   services like printers that are physically nearby, but the user is
   still free to browse any Zone in the offered list that they wish.

   An Epson printer may be preconfigured at the factory with the Service
   Name "Epson Printer", but they do not know on which network the
   printer will eventually be installed, so the printer will have to
   learn this from the network on arrival. On a network with multiple
   AppleTalk Zones, the AppleTalk network infrastructure informs the
   client of a single default Zone within which it may register Service
   Names. In the case of a device with a human user, the AppleTalk
   network infrastructure may also inform the client of a list of Zones
   within which the client may register Service Names, and the user may
   choose to register Service Names in any one of those Zones instead of
   in the suggested default Zone.

   Any NBP replacement needs to provide the following information to
   the client:
   
   * The suggested zone (domain) in which to register Service Names.
   * A list of recommended available zones (domains) in which Service
     Names may be optionally registered.
   * The suggested default zone (domain) for network browsing.
   * A list of available zones (domains) which may be browsed.

---------

Stuart Cheshire <cheshire@apple.com>
 * Wizard Without Portfolio, Apple Computer
 * Chairman, IETF ZEROCONF
 * www.stuartcheshire.org




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 17:12:32 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA15249
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 17:12:30 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ifUe-0006Zy-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 10:14:08 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ifUd-0006Zs-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 10:14:07 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ifUd-000D8y-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 10:14:07 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Sun, 16 Sep 2001 08:08:38 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Stuart Cheshire <cheshire@apple.com>
cc: Matt Crawford <crawdad@fnal.gov>, namedroppers@ops.ietf.org
Subject: Re: Question on Stewart's last mDNS item
In-Reply-To: <200109160822.BAA11508@scv3.apple.com>
Message-ID: <Pine.BSF.4.21.0109160748210.7834-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> >Did svrloc crash & burn without me noticing?  Or is there a
> >terrible gap in coverage?
> 
> A while back I promised you an answer to this question. It took me a 
> while, but here it is.
> 
> I realized that one goal of Multicast DNS was to replace AppleTalk NBP,

Indeed, replacement of legacy protocols (including Novell IPX and
NetBIOS) was one of the major goals discussed at the original
Zeroconf BOF. I also believe that this goal has been at the root of some
of the major arguments we've had about the overall direction of the mDNS
effort. 

Unfortunately, until now this goal was never articulated, and therefore
did not find its way into the Zeroconf requirements document. So in
writing the mDNS draft, we assumed that the Zeroconf WG had decided that
it wasn't important.

> but no one had actually written down precisely what is required to 
> replace AppleTalk NBP. 

If, as you suggest, we have been working on the solution to a problem
whose requirements have not yet been defined, then I believe that this
represents a major deficiency in the Zeroconf requirements document. 
Section 2.2.1 of draft-ietf-zeroconf-reqts-09.txt mentions that:

"A mechanism to support DNS resolver interfaces in the zero configuration
environment is required. 

Requirement:
- MUST support DNS application layer interfaces as described in RFC 1123,
  section 6.1 [RFC 1123]."

Section 2.2.2 mentions that:

"Requirement:

- MUST allow a host to determine if its name is unique. Then not
unique, notify the user or configuration software so that another name may
be chosen and similarly verified."

That's essentially it for the zeroconf name resolution requirements. 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 17:13:30 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA15262
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 17:13:30 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ifVe-0006bC-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 10:15:10 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ifVe-0006b6-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 10:15:10 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ifVd-000DAd-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 10:15:09 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109161551.f8GFpfH42517@as.vix.com>
To: namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language 
In-Reply-To: Message from Stuart Cheshire <cheshire@apple.com> 
   of "Sat, 15 Sep 2001 23:27:12 PDT." <200109160627.XAA05924@scv3.apple.com> 
Date: Sun, 16 Sep 2001 08:51:41 -0700
From: Paul A Vixie <vixie@vix.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Version 2 is useful for a home user who simply wants to name the devices 
> on his or her home network, without the expense or hassle of having to 
> pay some registry to "buy" some chunk of the global DNS name space within 
> which he or she can allocate names. This is much like what users do today 
> with AppleTalk NBP names, or Microsoft NetBIOS names.

The example which convinced me of the need for mDNS was when Woody said "what
about when two people meet in a bus terminal and want to do iRDA between Palm
devices?"  It is compellingly impossible to imagine DHCP being useful for this.

Folks, this isn't an ad-hoc hack.  mDNS fills a necessary gap in the IP stack.
If the consensus of the WG is otherwise, then the work will just go on outside
the IETF.  This could mean Palms never speak IP over their iRDA's, or it could
mean that mDNS is yet another socially unacceptable "experimental" protocol,
but there's no way that this functionality won't be present.  There's only a
way that IETF can disassociate itself from the functionality.


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 17:13:47 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA15272
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 17:13:47 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15igdK-0007SJ-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 11:27:10 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15igdK-0007SD-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 11:27:10 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15igdK-000F4e-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 11:27:10 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Sun, 16 Sep 2001 11:12:31 -0700 (PDT)
From: Bill Woodcock <woody@zocalo.net>
To: Stuart Cheshire <cheshire@apple.com>
cc: <namedroppers@ops.ietf.org>
Subject: Re: New mDNS configuration language
In-Reply-To: <200109160627.XAA05924@scv3.apple.com>
Message-ID: <Pine.NEB.4.33.0109161107520.6188-100000@woody.zocalo.net>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

    > 1. Use of IP multicast to look up DNS RRs that have names in the global
    > DNS name space

In other words, substitution of multicast for unicast with the overlying
DNS protocol and method of use otherwise unchanged.  Swap at layer 3.
Other things _above layer 5_ may query or use the information differently,
if they know they can get it via multicast, but that doesn't affect layer
5.

    > 2. Use of IP multicast to look up DNS RRs that have names within a
    > portion of the DNS name space specially designated to have only local
    > significance, akin to Net 10 IPv4 addresses, or 169.254/16 link-local
    > IPv4 addresses.

I've never been able to convince myself to care one whit about this
subset.  But I guess there are folks who find it interesting, and really
relish the idea of typing ".local" a lot.  That's great for them, as long
as they don't make me do it too.

			-Bill



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sun Sep 16 19:04:54 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA16055
	for <dnsext-archive@lists.ietf.org>; Sun, 16 Sep 2001 19:04:53 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15ikWq-000Bab-00
	for namedroppers-data@psg.com; Sun, 16 Sep 2001 15:36:44 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15ikWp-000BaV-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 15:36:43 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15ikWp-000Lck-00
	for namedroppers@ops.ietf.org; Sun, 16 Sep 2001 15:36:43 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>
Message-Id: <200109162152.GAA04672@necom830.hpcl.titech.ac.jp>
Received: by necom830.hpcl.titech.ac.jp (8.6.11/TM2.1)
	id GAA04672; Mon, 17 Sep 2001 06:52:18 +0900
Subject: Re: New mDNS configuration language
In-Reply-To: <200109161551.f8GFpfH42517@as.vix.com> from Paul A Vixie at "Sep
 16, 2001 08:51:41 am"
To: Paul A Vixie <vixie@vix.com>
Date: Mon, 17 Sep 2001 06:52:18 +0859 ()
CC: namedroppers@ops.ietf.org
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Paul;

> > Version 2 is useful for a home user who simply wants to name the devices 
> > on his or her home network, without the expense or hassle of having to 
> > pay some registry to "buy" some chunk of the global DNS name space within 
> > which he or she can allocate names. This is much like what users do today 
> > with AppleTalk NBP names, or Microsoft NetBIOS names.
> 
> The example which convinced me of the need for mDNS was when Woody said "what
> about when two people meet in a bus terminal and want to do iRDA between Palm
> devices?"  It is compellingly impossible to imagine DHCP being useful for this.

The example is called "ad-hoc networking".

The example has nothing to do with the Internet.

> Folks, this isn't an ad-hoc hack.

It is ad-hoc networking.

An examle related to the internetworking is two people meet in a bus
terminal with their Palm devices connected to the Internet through
a wireless base station at the bus terminal.

> mDNS fills a necessary gap in the IP stack.

The basic problem is that "stateless autoconfiguration" is yet
another hype of IPv6 and is impossible to archieve in an open
environment such as the Internet.

							Masataka Ohta

PS

An important question would be "when will the wireless internet service
become widely available?".

So, just FYI, I have actually started up a company to commercially offer
mobile IP service with wireless LAN this April. ;-)

	http://www.miserv.net/miserv/english/news/index.html


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 17 10:21:17 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA12976
	for <dnsext-archive@lists.ietf.org>; Mon, 17 Sep 2001 10:21:16 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15iynS-000Nte-00
	for namedroppers-data@psg.com; Mon, 17 Sep 2001 06:50:50 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15iynS-000NtY-00
	for namedroppers@ops.ietf.org; Mon, 17 Sep 2001 06:50:50 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15iynS-000Fdz-00
	for namedroppers@ops.ietf.org; Mon, 17 Sep 2001 06:50:50 -0700
Message-Id: <200109171059.GAA07670@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce: ;
Cc: namedroppers@ops.ietf.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-dnsext-mdns-04.txt
Date: Mon, 17 Sep 2001 06:59:49 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the DNS Extensions Working Group of the IETF.

	Title		: Multicast DNS
	Author(s)	: L. Esibov, B. Aboba, D. Thaler
	Filename	: draft-ietf-dnsext-mdns-04.txt
	Pages		: 17
	Date		: 14-Sep-01
	
Today, with the rise of home networking, there are an increasing number
of ad-hoc networks operating without a DNS server. In order to allow DNS
name resolution in such environments, the use of a multicast DNS is
proposed.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-dnsext-mdns-04.txt

To remove yourself from the IETF Announcement list, send a message to 
ietf-announce-request with the word unsubscribe in the body of the message.

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-dnsext-mdns-04.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-dnsext-mdns-04.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20010914103608.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-dnsext-mdns-04.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-dnsext-mdns-04.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20010914103608.I-D@ietf.org>

--OtherAccess--

--NextPart--




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 17 11:32:43 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA14024
	for <dnsext-archive@lists.ietf.org>; Mon, 17 Sep 2001 11:32:41 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15izyB-000OoZ-00
	for namedroppers-data@psg.com; Mon, 17 Sep 2001 08:05:59 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15izyB-000OoT-00
	for namedroppers@ops.ietf.org; Mon, 17 Sep 2001 08:05:59 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15izyB-000Hob-00
	for namedroppers@ops.ietf.org; Mon, 17 Sep 2001 08:05:59 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Mon, 17 Sep 2001 07:56:49 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: namedroppers@ops.ietf.org
Subject: Re: revised mDNS configuration language
Message-ID: <Pine.BSF.4.21.0109170753360.9314-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Erik Guttman spake thusly:

>
> Bernard,
> 
> I support this new text. 
> 
> Aidan,
> 
> This changes the focus of the 2937 revision.  The idea is to separate
> the question of 'whether?' from 'which order?'  The reason for this is
> the search order can't say NO, it only says which order (from which you
> imply YES).
>
> Also, the mDNS enable/disable option provides per interface
> configuration while RFC 2937 provides global configuration.
 
Here is another shot at the text for section 3.1, that tries to take
this into account:

3.1.  mDNS configuration

Multicast DNS usage can be configured manually or automatically.  
On interfaces where no manual or automatic configuration has been 
performed, multicast DNS is enabled by default.

For IPv6, the stateless DNS discovery mechanisms described in [19] can
be used to discover whether multicast DNS is globally enabled or
disabled.

Where DHCPv4 or DHCPv6 is implemented, DHCP options can be used to
configure multicast DNS on an interface. The mDNS enable DHCP option,
described in [6], can be used to explicitly enable or disable use of
multicast DNS on an interface. The Name Service Search option, described
in RFC 2937 [3], can be used to globally configure where multicast DNS is
used within the name service search order. DHCP option codes are used as
RFC 2937 codes signifying name services within the search order. As a
result, to specify multicast DNS usage within the name service search
order, the option code assigned to the mDNS enable option is used.

If an interface has been configured via any automatic configuration
mechanism which is able to supply DNS configuration information, then
multicast DNS MUST NOT be used on that interface unless it has been
explicitly enabled, whether via that mechanism or any other. This
ensures that upgraded hosts do not change their default behavior,
without requiring the source of the configuration information to be
simultaneously updated.  This implies that on the interface, the host
will neither listen on the DNS LINKLOCAL multicast address, nor will it
send queries to that address.  For a DNS server, automatic configuration
mechanisms MUST NOT  enable multicast DNS on any interface.



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 17 22:05:24 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA27911
	for <dnsext-archive@lists.ietf.org>; Mon, 17 Sep 2001 22:05:23 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15j9dD-0005zh-00
	for namedroppers-data@psg.com; Mon, 17 Sep 2001 18:24:59 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15j9dC-0005zb-00
	for namedroppers@ops.ietf.org; Mon, 17 Sep 2001 18:24:58 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15j9dC-0008p3-00
	for namedroppers@ops.ietf.org; Mon, 17 Sep 2001 18:24:58 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Robert Elz <kre@munnari.OZ.AU>
To: namedroppers@ops.ietf.org
Subject: Re: revised mDNS configuration language 
Date: Tue, 18 Sep 2001 08:13:05 +0700
Message-ID: <2789.1000775585@brandenburg.cs.mu.OZ.AU>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

There's another method that could be used for deciding when to use
mDNS (in the presence of some kind of configuration of course, the
decision when there has been no configuration at all has to have
been made by the software vendor - and while the suggestion for what
vendors should do in that case might be an interesting issue, nothing
else about it is).

When there's any configuration, which is always needed (currently
anyway) to do regular DNS, the one thing that must be told to the
end-nodes is where to find the back end resolver.

mDNS is essentially a method to allow that configuration to be
avoided, by defining a static address that vendors can install,
rather than requiring use of a particular local address that
necessarily has to be configured locally.    That mDNS also
allows nodes to be their own servers isn't relevant to this
message.

Earlier we discussed the possibility of simply having mDNS on
when there was no other DNS configuration, and off when there was,
but that didn't seem flexible enough (sites might want both in
some cases, with the lookups done in a particular order).

I know this is a bad idea, but I wanted to get it into the archives
anyway, so it will be there for future reference ... what could be
done, would be to simply allow the mDNS multicast address to be
configured as one of the local back end resolver addresses, if mDNS
is to be enabled.

That is, if the DNS server (really back end resolver) address configuration
is 192.168.1.9 224.0.0.99 172.29.28.27 (and assuming 224.0.0.99 was a
rational multicast address for mDNS to use) then that would be saying
to try 192.168.1.9 first, using DNS, then use mDNS, then try ...

The multicast address configured would just be the "base" address, in
which the part formed from the name hash (assuming that's how the protocol
ends up) would be included.   But it would give the multicast scope to use.

If we were starting from scratch, I can almost convince myself that this
would be  workable approach, and much simpler than anything else suggested.

Of course, current DNS configuration mechanisms supply data to current
DNS clients, which don't have a clue about use of mDNS, so actually doing
this would most likely cause far more problems than it could ever solve.

So, now everyone can just ignore this message...

kre



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 11:43:35 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA23353
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 11:43:35 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jLrG-000FQx-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 07:28:18 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jLrF-000FQr-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 07:28:17 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jLrF-0003eZ-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 07:28:17 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Tue, 18 Sep 2001 15:42:21 +0200 (MEST)
Message-Id: <20010918.154221.94131268.liman@autonomica.se>
To: namedroppers@ops.ietf.org
Subject: NCACHE mandatory for DS?
From: Lars-Johan Liman <liman@autonomica.se>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Something occured to me, while reading section 2.1.1 of =D3lafur's
draft-ietf-dnsext-delegation-signer-01.txt. He mentions the potential
problems of NS and DS having different TTLs, and thereby causing extra
queries for (timed out) DS records.

Please, help me think here. Won't this be the general case, so to
speak?

What if there is no DS RR for the child zone (it's insecure!)? Then a
DS-aware resolver will go look for one. It doesn't find it. Oh-well,
insecure child zone - no big deal. Next time around it will find the
NS RR in the cache, but not the DS RR (not because of TTL timeout, but
because of non-existence!) - so it goes looking for one, doesn't find
it, oh-well, insecure child zone - no big deal. Next time around -
same thing, and again, and again.

This is of course strongly mittigated by negative cacheing, but if you
_don't_ have that, you're going to have a steady stream of DS queries
for every normal query to an insecure zone, right? Or is there some
other mechanism, that I've misplaced in my memory, that kicks in to
prevent this behaviour? There is no NULL key to cache ...

Is there reason to make NCACHE mandatory in DS-aware resolvers? Is
there even reason to recommend icky tweaking of the negative TTL for
non-existing DS RRs, to match the corresponding NS RRs?

				Cheers,
				  /Liman
#----------------------------------------------------------------------
# Lars-Johan Liman		! E-mail: liman@autonomica.se
# Senior Systems Specialist     ! HTTP  : //www.autonomica.se/
# Autonomica AB, Stockholm 	! Voice : +46 8 - 615 85 72
#----------------------------------------------------------------------


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 13:13:01 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA26361
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 13:12:59 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jNZz-000H5P-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 09:18:35 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jNZz-000H5J-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 09:18:35 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jNZz-0006kb-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 09:18:35 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: Lars-Johan Liman <liman@autonomica.se>
Cc: namedroppers@ops.ietf.org
Subject: Re: NCACHE mandatory for DS?
References: <20010918.154221.94131268.liman@autonomica.se>
From: Derek Atkins <warlord@MIT.EDU>
Date: 18 Sep 2001 12:14:17 -0400
In-Reply-To: Lars-Johan Liman's message of "Tue, 18 Sep 2001 15:42:21 +0200 (MEST)"
Message-ID: <sjmofo8jx2e.fsf@rcn.ihtfp.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Wouldn't the NXT record (or otherwise) show that the DS record should
not exist anyways?  Or is the resolver not smart enough to notice that
before requested a non-existent/non-cached record?

-derek

PS: (I suppose this just pushes the onus onto the NXT record instead
of the DS record, or the SEC record, for insecure zones).

Lars-Johan Liman <liman@autonomica.se> writes:

> Something occured to me, while reading section 2.1.1 of =D3lafur's
> draft-ietf-dnsext-delegation-signer-01.txt. He mentions the potential
> problems of NS and DS having different TTLs, and thereby causing extra
> queries for (timed out) DS records.
> 
> Please, help me think here. Won't this be the general case, so to
> speak?
> 
> What if there is no DS RR for the child zone (it's insecure!)? Then a
> DS-aware resolver will go look for one. It doesn't find it. Oh-well,
> insecure child zone - no big deal. Next time around it will find the
> NS RR in the cache, but not the DS RR (not because of TTL timeout, but
> because of non-existence!) - so it goes looking for one, doesn't find
> it, oh-well, insecure child zone - no big deal. Next time around -
> same thing, and again, and again.
> 
> This is of course strongly mittigated by negative cacheing, but if you
> _don't_ have that, you're going to have a steady stream of DS queries
> for every normal query to an insecure zone, right? Or is there some
> other mechanism, that I've misplaced in my memory, that kicks in to
> prevent this behaviour? There is no NULL key to cache ...
> 
> Is there reason to make NCACHE mandatory in DS-aware resolvers? Is
> there even reason to recommend icky tweaking of the negative TTL for
> non-existing DS RRs, to match the corresponding NS RRs?
> 
> 				Cheers,
> 				  /Liman
> #----------------------------------------------------------------------
> # Lars-Johan Liman		! E-mail: liman@autonomica.se
> # Senior Systems Specialist     ! HTTP  : //www.autonomica.se/
> # Autonomica AB, Stockholm 	! Voice : +46 8 - 615 85 72
> #----------------------------------------------------------------------
> 
> 
> to unsubscribe send a message to namedroppers-request@ops.ietf.org with
> the word 'unsubscribe' in a single line as the message text body.

-- 
       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord@MIT.EDU                        PGP key available


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 14:18:26 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA28267
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 14:18:26 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jP6m-000InO-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 10:56:32 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jP6m-000InH-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 10:56:32 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jP6l-0009Rr-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 10:56:31 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Tue, 18 Sep 2001 19:51:12 +0200 (CEST)
From: Roy Arends <Roy.Arends@nominum.com>
To: Derek Atkins <warlord@MIT.EDU>
Cc: Lars-Johan Liman <liman@autonomica.se>, <namedroppers@ops.ietf.org>
Subject: Re: NCACHE mandatory for DS?
In-Reply-To: <Pine.BSF.4.33.0109181940180.310-100000@node10c4d.a2000.nl>
Message-ID: <Pine.BSF.4.33.0109181945380.310-100000@node10c4d.a2000.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Whoops, replying to myself.

On Tue, 18 Sep 2001, Roy Arends wrote:

> On 18 Sep 2001, Derek Atkins wrote:
>
> > Wouldn't the NXT record (or otherwise) show that the DS record should
> > not exist anyways?  Or is the resolver not smart enough to notice that
> > before requested a non-existent/non-cached record?
>
> When the initial query-response is cached, it has either the set NS,DS,SIG
> cached or the set NXT SIG cached for the qname and qtype. If the first set
> is cached, and the DS expires, there is no other way then specifically
> querying for DS, since it has no NXT cached.

To be complete: the NS,DS,SIG set is only when the child is secured. When
the child is unsecure, the set NS,NXT,SIG will do the trick, since NXT
signals the absence of DS. This is also mentioned in the draft.

Regards,

Roy Arends
Nominum






to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 14:18:29 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA28279
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 14:18:28 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jP0C-000Igc-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 10:49:44 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jP0C-000IgW-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 10:49:44 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jP0C-0009GX-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 10:49:44 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Tue, 18 Sep 2001 19:44:23 +0200 (CEST)
From: Roy Arends <Roy.Arends@nominum.com>
To: Derek Atkins <warlord@MIT.EDU>
Cc: Lars-Johan Liman <liman@autonomica.se>, <namedroppers@ops.ietf.org>
Subject: Re: NCACHE mandatory for DS?
In-Reply-To: <sjmofo8jx2e.fsf@rcn.ihtfp.org>
Message-ID: <Pine.BSF.4.33.0109181940180.310-100000@node10c4d.a2000.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On 18 Sep 2001, Derek Atkins wrote:

> Wouldn't the NXT record (or otherwise) show that the DS record should
> not exist anyways?  Or is the resolver not smart enough to notice that
> before requested a non-existent/non-cached record?

When the initial query-response is cached, it has either the set NS,DS,SIG
cached or the set NXT SIG cached for the qname and qtype. If the first set
is cached, and the DS expires, there is no other way then specifically
querying for DS, since it has no NXT cached.

Regards,

Roy

> Lars-Johan Liman <liman@autonomica.se> writes:
>
> > Something occured to me, while reading section 2.1.1 of =D3lafur's
> > draft-ietf-dnsext-delegation-signer-01.txt. He mentions the potential
> > problems of NS and DS having different TTLs, and thereby causing extra
> > queries for (timed out) DS records.
> >
> > Please, help me think here. Won't this be the general case, so to
> > speak?
> >
> > What if there is no DS RR for the child zone (it's insecure!)? Then a
> > DS-aware resolver will go look for one. It doesn't find it. Oh-well,
> > insecure child zone - no big deal. Next time around it will find the
> > NS RR in the cache, but not the DS RR (not because of TTL timeout, but
> > because of non-existence!) - so it goes looking for one, doesn't find
> > it, oh-well, insecure child zone - no big deal. Next time around -
> > same thing, and again, and again.
> >
> > This is of course strongly mittigated by negative cacheing, but if you
> > _don't_ have that, you're going to have a steady stream of DS queries
> > for every normal query to an insecure zone, right? Or is there some
> > other mechanism, that I've misplaced in my memory, that kicks in to
> > prevent this behaviour? There is no NULL key to cache ...
> >
> > Is there reason to make NCACHE mandatory in DS-aware resolvers? Is
> > there even reason to recommend icky tweaking of the negative TTL for
> > non-existing DS RRs, to match the corresponding NS RRs?
> >
> > 				Cheers,
> > 				  /Liman
> > #----------------------------------------------------------------------
> > # Lars-Johan Liman		! E-mail: liman@autonomica.se
> > # Senior Systems Specialist     ! HTTP  : //www.autonomica.se/
> > # Autonomica AB, Stockholm 	! Voice : +46 8 - 615 85 72
> > #----------------------------------------------------------------------
> >
> >
> > to unsubscribe send a message to namedroppers-request@ops.ietf.org with
> > the word 'unsubscribe' in a single line as the message text body.
>
> --
>        Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
>        Member, MIT Student Information Processing Board  (SIPB)
>        URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
>        warlord@MIT.EDU                        PGP key available
>
>
> to unsubscribe send a message to namedroppers-request@ops.ietf.org with
> the word 'unsubscribe' in a single line as the message text body.
>

-------------
0-14-023750-X dcrpt ths 43.0D.01 01.05.0C 84.18.03 8A.13.04 2D.0B.0A




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 14:23:45 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA28376
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 14:23:44 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jP6u-000Ing-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 10:56:40 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jP6u-000Ina-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 10:56:40 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jP6u-0009SA-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 10:56:40 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: Lars-Johan Liman <liman@autonomica.se>
Cc: namedroppers@ops.ietf.org
Subject: Re: NCACHE mandatory for DS? 
In-Reply-To: Your message of "Tue, 18 Sep 2001 15:42:21 +0200."
             <20010918.154221.94131268.liman@autonomica.se> 
Date: Tue, 18 Sep 2001 10:53:11 -0700
Message-ID: <39570.1000835591@shell.nominum.com>
From: Jim Reid <Jim.Reid@nominum.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>>>>> "Liman" == Lars-Johan Liman <liman@autonomica.se> writes:

    Liman> Is there reason to make NCACHE mandatory in DS-aware
    Liman> resolvers?

Perhaps. But since the DNS implementations which will presumably
implement DS already support negative caching, the question is moot.



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 21:01:26 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id VAA05694
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 21:01:25 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jVEa-0000T5-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 17:29:00 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jVEZ-0000Sz-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 17:28:59 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jVEZ-000MtX-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 17:28:59 -0700
From: Randy Bush <randy@psg.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: namedroppers <namedroppers@ops.ietf.org>
Subject: [e2e] DNS
Message-Id: <E15jVEZ-000MtX-00@rip.psg.com>
Date: Tue, 18 Sep 2001 17:28:59 -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

[ excuse the crosspost, but ... ]

[ the caching result for a rrs is good.  i am aware of other research which
  shows caching of ns rrs is quite beneficial -- rb ]

---


From: Hari Balakrishnan <hari@lcs.mit.edu>
To: end2end-interest@postel.org
Subject: [e2e] DNS
Date: Tue, 18 Sep 2001 20:02:21 -0400


We've been collecting client-side traffic traces of DNS traffic, jointly with 
the driving TCP connection workload, for a couple of years now, and a recent 
paper analyzing these data sets may be of interest to many on this list.  Our 
data spans many months of collection at the border router connecting MIT's LCS 
and AI Lab to the rest of the Internet, as well as a border router at KAIST in 
Korea.

The paper, "DNS Performance and the Effectiveness of Caching," 
by Jaeyeon Jung, Emil Sit, Hari Balakrishnan, and Robert Morris, is at 
http://nms.lcs.mit.edu/papers/dns-imw2001.html

Some of our non-obvious (to us!) findings and conclusions are: 

- About a quarter of all DNS lookups never get an answer. More than 50% of
the DNS-related packets in the wide-area correspond to such lookups...

- The DNS retransmission protocol appeears to be overly persistent: while
most successful answers are received in at most 2-3 retransmissions, the
lack of an answer or response causes a much larger number of retransmissions
and a corresponding number of DNS packets traverse the wide-area.

- Replacing the A-record TTL's for most (or all) Internet hosts to a value
as small of 10 minutes is not likely to degrade the scalability of DNS in
any noticeable way.

- The scalability of DNS has little to do with its hierarchical organization
or good A-record caching. Most of the DNS name space is a flat, two-level
structure.  A-record caching does not seem to add much more to the per-host
or per-application caching done by end clients today.  Rather, the
scalability derives from the good name space partitioning achieved by the
cacheability of NS records, which avoid load on the root and top-level name
servers.

The paper details the combination of trace analysis and trace-driven cache
simulation that we used to arrive at these conclusions.

to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 23:39:06 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id XAA09420
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 23:39:06 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jXyQ-0003Pu-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 20:24:30 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jXyP-0003Po-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 20:24:29 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jXyP-0001pF-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 20:24:29 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Randy Bush <randy@psg.com>
To: Bernard Aboba <aboba@internaut.com>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: [e2e] DNS
References: <E15jVEZ-000MtX-00@rip.psg.com>
	<Pine.BSF.4.21.0109182013320.11536-100000@internaut.com>
Message-Id: <E15jXy4-0001oa-00@rip.psg.com>
Date: Tue, 18 Sep 2001 20:24:08 -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>> [ the caching result for a rrs is good.  i am aware of other research
>>   which shows caching of ns rrs is quite beneficial -- rb ]
> The packet loss result is not so good -- 25%! Perhaps there is a case for
> standardizing resolver retransmission behavior after all.

oh, indeed!  my comment was merely to limit any undue inferences that might
be made from the a rr caching result.

randy


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 18 23:42:05 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id XAA09464
	for <dnsext-archive@lists.ietf.org>; Tue, 18 Sep 2001 23:42:04 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jXx9-0003NB-00
	for namedroppers-data@psg.com; Tue, 18 Sep 2001 20:23:11 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jXx8-0003N5-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 20:23:10 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jXx8-0001ms-00
	for namedroppers@ops.ietf.org; Tue, 18 Sep 2001 20:23:10 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Tue, 18 Sep 2001 20:14:57 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Randy Bush <randy@psg.com>
cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: [e2e] DNS
In-Reply-To: <E15jVEZ-000MtX-00@rip.psg.com>
Message-ID: <Pine.BSF.4.21.0109182013320.11536-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> [ the caching result for a rrs is good.  i am aware of other research which
>   shows caching of ns rrs is quite beneficial -- rb ]

The packet loss result is not so good -- 25%! Perhaps there is a case for
standardizing resolver retransmission behavior after all. 




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 19 03:56:13 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id DAA26323
	for <dnsext-archive@lists.ietf.org>; Wed, 19 Sep 2001 03:56:13 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15jbq0-0007v3-00
	for namedroppers-data@psg.com; Wed, 19 Sep 2001 00:32:04 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15jbpz-0007ux-00
	for namedroppers@ops.ietf.org; Wed, 19 Sep 2001 00:32:03 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15jbpz-0008pK-00
	for namedroppers@ops.ietf.org; Wed, 19 Sep 2001 00:32:03 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BA845AF.79D82209@email.mot.com>
Date: Wed, 19 Sep 2001 17:13:51 +1000
From: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
To: Robert Elz <kre@munnari.OZ.AU>
CC: namedroppers@ops.ietf.org
Subject: Re: revised mDNS configuration language
References: <2789.1000775585@brandenburg.cs.mu.OZ.AU>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Robert Elz wrote:
> That is, if the DNS server (really back end resolver) address configuration
> is 192.168.1.9 224.0.0.99 172.29.28.27 (and assuming 224.0.0.99 was a
> rational multicast address for mDNS to use) then that would be saying
> to try 192.168.1.9 first, using DNS, then use mDNS, then try ...
> 

> Of course, current DNS configuration mechanisms supply data to current
> DNS clients, which don't have a clue about use of mDNS, so actually doing
> this would most likely cause far more problems than it could ever solve.
> 

Such as?

  - There isn't anything I know of *requiring* those back-end resolvers
    to be consulted in the order that the addresses are provided.

  - If you got a "name does not exist" from a DNS server,
    the resolver would not bother to try the next address
    (ie the mDNS one)

  - DNS over TCP would be broken.

Looks pretty infeasible to me.

> So, now everyone can just ignore this message...
> 

I'm trying, but I know a couple of people who have tried this,
and it seemed to work OK.  Of course, "works" is a relative term.

regards
	aidan
____
:wq!

Communications and Networks Lab         aidan.williams@motorola.com
Motorola Australian Research Centre          phone: +61 2 9666 0649
Locked Bag 5028, Botany NSW 1455              phax: +61 2 9666 0501
Australia                            http://cnlab.arc.corp.mot.com/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 19 18:21:35 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA11827
	for <dnsext-archive@lists.ietf.org>; Wed, 19 Sep 2001 18:21:35 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15josX-000BDx-00
	for namedroppers-data@psg.com; Wed, 19 Sep 2001 14:27:33 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15josX-000BDr-00
	for namedroppers@ops.ietf.org; Wed, 19 Sep 2001 14:27:33 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15josX-000MA0-00
	for namedroppers@ops.ietf.org; Wed, 19 Sep 2001 14:27:33 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BA90C80.CEDA80F6@daimlerchrysler.com>
Date: Wed, 19 Sep 2001 17:22:08 -0400
From: Kevin Darcy <kcd@daimlerchrysler.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: [e2e] DNS
References: <E15jVEZ-000MtX-00@rip.psg.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

While the report was a bit of an eye-opener, I think the conclusion that
(paraphrased) "reducing A-record TTLs won't hurt" is a little facile,
especially given the other conclusion that (paraphrasing again) "NS-record
caching is saving our collective butts". After all, the referral-following
mechanism requires A-records *in*addition* to NS-records and if the TTLs on
*those* A-records drop precipitously, then most of the benefits of NS
caching will be lost. The authors of the paper seem to be treating
NS-records as if they exist entirely separately from A-records, but in
reality of course NS-records and A-records and the respective caching
characteristics of each are interdependent as the protocol stands today. It
would take something like my "REFER" record-type proposal (see the archives)
to reduce or eliminate that interdependency.



- Kevin

Randy Bush wrote:

> [ excuse the crosspost, but ... ]
>
> [ the caching result for a rrs is good.  i am aware of other research which
>   shows caching of ns rrs is quite beneficial -- rb ]
>
> ---
>
> From: Hari Balakrishnan <hari@lcs.mit.edu>
> To: end2end-interest@postel.org
> Subject: [e2e] DNS
> Date: Tue, 18 Sep 2001 20:02:21 -0400
>
> We've been collecting client-side traffic traces of DNS traffic, jointly with
> the driving TCP connection workload, for a couple of years now, and a recent
> paper analyzing these data sets may be of interest to many on this list.  Our
> data spans many months of collection at the border router connecting MIT's LCS
> and AI Lab to the rest of the Internet, as well as a border router at KAIST in
> Korea.
>
> The paper, "DNS Performance and the Effectiveness of Caching,"
> by Jaeyeon Jung, Emil Sit, Hari Balakrishnan, and Robert Morris, is at
> http://nms.lcs.mit.edu/papers/dns-imw2001.html
>
> Some of our non-obvious (to us!) findings and conclusions are:
>
> - About a quarter of all DNS lookups never get an answer. More than 50% of
> the DNS-related packets in the wide-area correspond to such lookups...
>
> - The DNS retransmission protocol appeears to be overly persistent: while
> most successful answers are received in at most 2-3 retransmissions, the
> lack of an answer or response causes a much larger number of retransmissions
> and a corresponding number of DNS packets traverse the wide-area.
>
> - Replacing the A-record TTL's for most (or all) Internet hosts to a value
> as small of 10 minutes is not likely to degrade the scalability of DNS in
> any noticeable way.
>
> - The scalability of DNS has little to do with its hierarchical organization
> or good A-record caching. Most of the DNS name space is a flat, two-level
> structure.  A-record caching does not seem to add much more to the per-host
> or per-application caching done by end clients today.  Rather, the
> scalability derives from the good name space partitioning achieved by the
> cacheability of NS records, which avoid load on the root and top-level name
> servers.
>
> The paper details the combination of trace analysis and trace-driven cache
> simulation that we used to arrive at these conclusions.





to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 20 10:17:47 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA11163
	for <dnsext-archive@lists.ietf.org>; Thu, 20 Sep 2001 10:17:45 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15k3zr-000Bnp-00
	for namedroppers-data@psg.com; Thu, 20 Sep 2001 06:36:07 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15k3zr-000Bnj-00
	for namedroppers@ops.ietf.org; Thu, 20 Sep 2001 06:36:07 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15k3zr-000Dkh-00
	for namedroppers@ops.ietf.org; Thu, 20 Sep 2001 06:36:07 -0700
Message-Id: <200109201106.HAA05195@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce: ;
Cc: namedroppers@ops.ietf.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-dnsext-delegation-signer-02.txt
Date: Thu, 20 Sep 2001 07:06:49 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the DNS Extensions Working Group of the IETF.

	Title		: Delegation Signer record in parent
	Author(s)	: O. Gudmundsson
	Filename	: draft-ietf-dnsext-delegation-signer-02.txt
	Pages		: 11
	Date		: 19-Sep-01
	
The Delegation Signer (DS) RR set is stored in a delegating (parent)
zone at each delegation point, and indicates the keys used in the
delegated (child) zone. The main design goal of the DS RR simplify the
operation of secure delegations by eliminating the need to store the
same RR in parent and child, as is done with the NS RR set and the KEY
set in RFC2535.
Secure resolvers need to take an additional step with DS to verify a
child's KEY RR set. Operationally this schema is much simpler as
operation of the two zones at delegation is now decoupled to great
extent.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-dnsext-delegation-signer-02.txt

To remove yourself from the IETF Announcement list, send a message to 
ietf-announce-request with the word unsubscribe in the body of the message.

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-dnsext-delegation-signer-02.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-dnsext-delegation-signer-02.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20010919121341.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-dnsext-delegation-signer-02.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-dnsext-delegation-signer-02.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20010919121341.I-D@ietf.org>

--OtherAccess--

--NextPart--




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 20 11:47:13 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA13837
	for <dnsext-archive@lists.ietf.org>; Thu, 20 Sep 2001 11:47:13 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15k5is-000EHe-00
	for namedroppers-data@psg.com; Thu, 20 Sep 2001 08:26:42 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15k5is-000EHY-00
	for namedroppers@ops.ietf.org; Thu, 20 Sep 2001 08:26:42 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15k5is-000GmM-00
	for namedroppers@ops.ietf.org; Thu, 20 Sep 2001 08:26:42 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 20 Sep 2001 17:20:40 +0200 (CEST)
From: Roy Arends <Roy.Arends@nominum.com>
To: <namedroppers@ops.ietf.org>
Subject: Re: I-D ACTION:draft-ietf-dnsext-delegation-signer-02.txt
In-Reply-To: <200109201106.HAA05195@ietf.org>
Message-ID: <Pine.BSF.4.33.0109201714070.312-100000@node10c4d.a2000.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

draft-ietf-dnsext-delegation-signer-02.txt

section 2.2.1

  Different TTL values on the child's NS set and parents DS set may cause
  the DS set to expire before the NS set. In this case an non-DS aware
  server would ask the child server for the DS set and get NXDOMAIN
  answer. DS aware server will know to ask a parent DNS server for the DS
  record.

Wouldn't that be a NODATA instead of NXDOMAIN ?

Regards,

Roy
-------------
0-14-023750-X dcrpt ths 43.0D.01 01.05.0C 84.18.03 8A.13.04 2D.0B.0A




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 20 15:07:42 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA19316
	for <dnsext-archive@lists.ietf.org>; Thu, 20 Sep 2001 15:07:41 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15k8iq-000JRb-00
	for namedroppers-data@psg.com; Thu, 20 Sep 2001 11:38:52 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15k8iq-000JRV-00
	for namedroppers@ops.ietf.org; Thu, 20 Sep 2001 11:38:52 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15k8ip-000MAP-00
	for namedroppers@ops.ietf.org; Thu, 20 Sep 2001 11:38:51 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 20 Sep 2001 11:30:33 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
cc: Erik Guttman <Erik.Guttman@sun.com>, namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
In-Reply-To: <3BA96F73.C8D9054E@email.mot.com>
Message-ID: <Pine.BSF.4.21.0109201126130.14278-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> I can't see a need to revise rfc2937 in light of this text.

Possibly this is not needed. 
> 
> However, if it is going to work this way, I think that the
> interactions between mdns-enable and the search order need to be
> spelled out (i.e. Erik's table probably need to be included
> somewhere), even if people don't bother to implement rfc2937.
> 

Erik - do you have some proposed text to insert?

> > If an interface has been configured via any automatic configuration
> > mechanism which is able to supply DNS configuration information, then
> > multicast DNS MUST NOT be used on that interface unless it has been
> > explicitly enabled, whether via that mechanism or any other. This
> > ensures that upgraded hosts do not change their default behavior,
> > without requiring the source of the configuration information to be
> > simultaneously updated.  This implies that on the interface, the host
> > will neither listen on the DNS LINKLOCAL multicast address, nor will it
> > send queries to that address.
> 
> I've been a system admin for a number of years, I know where this text
> is coming from, and I sympathise, but ...
> 
>   - DHCP == death-to-mdns, which is a somewhat stronger statement than
>     was being advocated on the list which was DNS-disables-mdns.
>     Your text is is certainly more consistent.
> 
>   - mdns will be disabled in environments where NetBIOS naming and
>     Appletalk naming operate adequately today.
>     mDNS as a replacement for those protocols becomes still-born.
> 
>     One such environment is the home.  I don't know of any NAT/DHCP
>     boxes that allow the user to add the mdns-enable option, even if
>     they could.  So then I have to enable it on every host.
> 

Yes, that is undesirable. But it's hard to figure out how not to mess up
the home case while giving enterprise admins what they want. Any ideas??

>     It exasperates me that the people least able to cope have to do
>     configuration, so that the people best able to cope (managers of
>     networks) don't have to do anything.

Yup. Any way to give both of them what they want?

> > For a DNS server, automatic configuration mechanisms MUST NOT enable
> > multicast DNS on any interface.
> 
> Why not?  I don't recall seeing a reason why recursive mDNS lookups
> should be disallowed.  If you aren't a back-end resolver, don't
> resolve them.  Am I missing something here?
> 

I think that there was a concern about nodes becoming mDNS relays. 

to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 21 07:40:30 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA19987
	for <dnsext-archive@lists.ietf.org>; Fri, 21 Sep 2001 07:40:28 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15kO5j-000M5b-00
	for namedroppers-data@psg.com; Fri, 21 Sep 2001 04:03:31 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15kO5i-000M5V-00
	for namedroppers@ops.ietf.org; Fri, 21 Sep 2001 04:03:30 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15kO5i-000CNZ-00
	for namedroppers@ops.ietf.org; Fri, 21 Sep 2001 04:03:30 -0700
Message-Id: <200109211102.HAA18952@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce: ;
Cc: namedroppers@ops.ietf.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-dnsext-mdns-05.txt
Date: Fri, 21 Sep 2001 07:02:30 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the DNS Extensions Working Group of the IETF.

	Title		: Multicast DNS
	Author(s)	: L. Esibov, B. Aboba, D. Thaler
	Filename	: draft-ietf-dnsext-mdns-05.txt
	Pages		: 17
	Date		: 20-Sep-01
	
Today, with the rise of home networking, there are an increasing number
of ad-hoc networks operating without a DNS server. In order to allow DNS
name resolution in such environments, the use of a multicast DNS is
proposed.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-dnsext-mdns-05.txt

To remove yourself from the IETF Announcement list, send a message to 
ietf-announce-request with the word unsubscribe in the body of the message.

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-dnsext-mdns-05.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-dnsext-mdns-05.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20010920115820.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-dnsext-mdns-05.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-dnsext-mdns-05.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20010920115820.I-D@ietf.org>

--OtherAccess--

--NextPart--




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 21 11:20:42 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA28254
	for <dnsext-archive@lists.ietf.org>; Fri, 21 Sep 2001 11:20:40 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15kRjp-0003aW-00
	for namedroppers-data@psg.com; Fri, 21 Sep 2001 07:57:09 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15kRjo-0003aQ-00
	for namedroppers@ops.ietf.org; Fri, 21 Sep 2001 07:57:08 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15kRjo-000Icm-00
	for namedroppers@ops.ietf.org; Fri, 21 Sep 2001 07:57:08 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Fri, 21 Sep 2001 15:28:18 +0200 (MET DST)
From: Erik Guttman <Erik.Guttman@Sun.COM>
Subject: Re: New mDNS configuration language
To: Bernard Aboba <aboba@internaut.com>
Cc: Aidan Williams <Aidan_Williams-A15677@email.mot.com>,
        Erik Guttman <Erik.Guttman@Sun.COM>, namedroppers@ops.ietf.org
In-Reply-To: "Your message with ID" <Pine.BSF.4.21.0109201126130.14278-100000@internaut.com>
Message-ID: <Roam.SIMC.2.0.6.1001078898.12056.erikg@ehdb03-home.germany>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Bernard, Aidan,

> > However, if it is going to work this way, I think that the
> > interactions between mdns-enable and the search order need to be
> > spelled out (i.e. Erik's table probably need to be included
> > somewhere), even if people don't bother to implement rfc2937.
> > 
> 
> Erik - do you have some proposed text to insert?

I suggest:

  The mDNS Enable Option specifies the order in which DNS and multicast
  DNS is done.  The mDNS Enable Option affects only DNS resolver behavior,
  that is, how DNS resolution is performed, whether multicast DNS is used.  

  The mDNS Enable Option does not determine whether or in which order DNS 
  itself is used for name resolution.  This can be specified, for example, 
  using the Name Service Search Option for DHCP [RFC2937].

> > > If an interface has been configured via any automatic configuration
> > > mechanism which is able to supply DNS configuration information, then
> > > multicast DNS MUST NOT be used on that interface unless it has been
> > > explicitly enabled, whether via that mechanism or any other. 

> >   - DHCP == death-to-mdns, which is a somewhat stronger statement than
> >     was being advocated on the list which was DNS-disables-mdns.

> >   - mdns will be disabled in environments where NetBIOS naming and
> >     Appletalk naming operate adequately today.
> >     mDNS as a replacement for those protocols becomes still-born.
> > 
> >     One such environment is the home.  I don't know of any NAT/DHCP
> >     boxes that allow the user to add the mdns-enable option, even if
> >     they could.  So then I have to enable it on every host.

> Yes, that is undesirable. But it's hard to figure out how not to mess up
> the home case while giving enterprise admins what they want. Any ideas??

Personally, I would prefer the policy 'allow unless explicitely disabled.'
Many resisted this proposal (for example, the WG chairs.)

If mDNS were another protocol (not DNS at all), we wouldn't have this
problem.  Then it, like appletalk and netbios, could coexist with DNS.
But we are defining a new behavior for DNS.  

Or should we use a different port #?  That would dissolve this problem.

> >     It exasperates me that the people least able to cope have to do
> >     configuration, so that the people best able to cope (managers of
> >     networks) don't have to do anything.
> 
> Yup. Any way to give both of them what they want?

I believe that once we have experience with mDNS in real networks folks
will become more comfortable with it.  We will be able to relax the
'MUST NOT enable by mDNS by default' to a 'do whatever' policy.  That's
my hope.

Erik



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 21 12:17:42 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA29899
	for <dnsext-archive@lists.ietf.org>; Fri, 21 Sep 2001 12:17:42 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15kSkJ-0005mj-00
	for namedroppers-data@psg.com; Fri, 21 Sep 2001 09:01:43 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15kSkI-0005mc-00
	for namedroppers@ops.ietf.org; Fri, 21 Sep 2001 09:01:42 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15kSkI-000KRo-00
	for namedroppers@ops.ietf.org; Fri, 21 Sep 2001 09:01:42 -0700
Date: Fri, 21 Sep 2001 08:48:18 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Erik Guttman <Erik.Guttman@Sun.COM>
cc: Aidan Williams <Aidan_Williams-A15677@email.mot.com>,
        namedroppers@ops.ietf.org
Subject: Re: New mDNS configuration language
In-Reply-To: <Roam.SIMC.2.0.6.1001078898.12056.erikg@ehdb03-home.germany>
Message-ID: <Pine.BSF.4.21.0109210846100.15945-100000@internaut.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

> Personally, I would prefer the policy 'allow unless explicitely disabled.'
> Many resisted this proposal (for example, the WG chairs.)

There is somewhat of a double standard. NetBIOS & AppleTalk are "allow
unless explicitly disabled" once installed. So if mDNS were to be used
instead with the same policy, it would fit within the same admin
model. Yet, admins I talk to seem to hate the idea of clients changing
their configuration in an upgrade. I think that is because they're afraid
that they will then be stuck with *both* mDNS & NetBIOS/AppleTalk ;)



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 24 10:52:10 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA15479
	for <dnsext-archive@lists.ietf.org>; Mon, 24 Sep 2001 10:52:09 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15lOzm-000Ark-00
	for namedroppers-data@psg.com; Sun, 23 Sep 2001 23:13:34 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15lOzl-000Are-00
	for namedroppers@ops.ietf.org; Sun, 23 Sep 2001 23:13:33 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15lOzl-0006Y4-00
	for namedroppers@ops.ietf.org; Sun, 23 Sep 2001 23:13:33 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BAECDE8.7DA9E8C0@email.mot.com>
Date: Mon, 24 Sep 2001 16:08:40 +1000
From: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
To: namedroppers@ops.ietf.org
Subject: Enabling mDNS in the home environment
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

The current approach to mdns configuration disables mdns when DHCP is
in operation, unless the mdns-enable DHCP option is present.  This has
the effect of disabling mdns by default in the place where I think
this proposal has the most to offer -- the home.

Enabling mdns by default is unacceptable to the working group.  What I
am looking for is a way of identifying the home, so that mdns may be
enabled as often as possible in that environment.  To that end there
are some possible proposals later, but to start with I'll describe
what I think the environment of "the home" is.


======================================================================
Characteristics of mDNS (slightly different to Stuart's two):

  - the "locally scoped namespace" .local.arpa.
    I believe this namespace is useful whether multicast
    transport is used or not.

  - multicast transport
    Hosts can answer DNS requests for names they know about,
    including .local.arpa. and some.fully.qualified.name.


======================================================================
Un-interesting "home":
  - single PC, no home network, no need for mdns, nothing else to say.

A more-interesting "home" has a network with multiple PCs and a
gateway connecting them to the internet.  The "home gateway box"
contains the following:

  - DHCP server, serving dynamic private IPv4 addresses
    (could be any rfc1918 prefix, but is typically 192.168/16)

    This DHCP server does not reflect the policy of the local
    "administrator", it is "pre-configured" at the "factory".
    ISP DNS server, domain name, search list usually reflected into
    the home via the DHCP server.

    If a DHCP server like this (which is *not* configured by an admin)
    could be identified, then mDNS could be enabled in this
    environment, which would be a good thing for the reasons given
    later.

  - Network Address Port Translation (NAPT)

  - optional DNS "back end resolver"/proxy
    This is reasonably rare, because the ISP DNS server can be
    consulted via the NAT.  One less piece of software in the box.

    I don't know of a home gateway with a DNS server supporting DDNS.

  - The whole thing is a black box.

    There is no way of adding a new DHCP option (such as mdns-enable)
    to the DHCP server configuration.

  - the picture:
                                 +--------+
     --------                    |  DNS   |
             \                   | Server |
     RFC1918 |   +-------+       +--------+
     Network |== |Home-GW| ==> ISP Network
     192.168 |   +-------+     Global IP addrs
            /                  Global name space
     -------

In such an environment, the .local.arpa. namespace is helpful for
naming devices in the home.  However the name service ends up being
provided, the intuitive thing to do is to configure the name into the
device.  Globally unique DNS names aren't required for every device in
the home, and even if you had them the chances of devices serving
hosts on the global internet through the NAT are pretty slim.  In this
environment, ip addresses are non-unique/locally scoped, and so it
makes sense to have non-unique/locally scoped names as well.


======================================================================
Ways of supporting the .local.arpa. namespace in the home:

  - DDNS to a DNS server in the gateway box

    Hosts could use DDNS directly, or the DHCP server in the gateway
    box could do it.  A DHCP server could use a supplied hostname to
    populate .local.arpa. some other way too.

    Requires a home gateway vendor to implement a DNS server, and
    integrate it with DHCP or DDNS.  Multiple DNS servers in the home
    desiring to manage .local.arpa. are tricky.  How do they
    co-operate?

    DDNS isn't mandatory in draft-aboba-dhc-mini-03.txt, which is the
    nearest thing I know of to a spec for this kind of home gateway.

  - DDNS to the ISP name server (or whereever)

    An ISP serving a bunch of these NAT boxes:
      * would need to allow many hosts do DDNS updates
        (I haven't found an ISP yet willing to do DDNS)
      * Name space shared amongst many people, name collision likely.
      * PTR wouldn't make much sense, same private address is potentially
        one of many names

    I question whether most ISPs have any desire to be involved in the
    management of names and addresses for every device in the home.

  - multicast transport for DNS requests in the home

    Hosts are named already in an intuitive way.  Get hosts to answer
    requests for their names.  Hosts don't have to rely on having a
    box on the network with the right collection of features.

    A gateway could assist non-mdns hosts by doing mdns lookups on
    behalf of clients who don't know how.

  - give up and stick with NetBIOS/Appletalk
    This might be the option we end up with if we can't sort something
    out for mdns.


To me, multicast transport for DNS names in conjunction with the
.local.arpa. namespace looks like a good solution for the home.


======================================================================
Enabling multicast DNS in the home environment,
or "how can we identify the home and enable mdns in it"

 a) Heuristic: I'm on a rfc1918 network, my DNS server isn't.

    Proposal: A host would do something like:

      if my_ip in 192.168/16 &&
	 dns_server_ip not in 192.168/16
      then
	# because we are natting to a DNS server in an ISP
        enable_mdns
      fi

    This isn't perfect.  Its a heuristic.  I deliberately picked the
    more restrictive 192.168/16 because 10/8 finds wide use in ISPs,
    corporate networks, etc.

 b) treat mDNS as a replacement for NetBIOS naming

    mDNS has many similarities to NetBIOS naming.  Broadcast NetBIOS
    name resolution is enabled by default.  In administered
    environments, broadcast name resolution is disabled by virtue of a
    WINS server address and B/P/M node configuration [rfc1001.txt].

    Proposal: that mdns be disabled when NetBIOS naming is configured,
	      enabled otherwise


There is limited information which is passed from the ISP, through the
DHCP proxy in the Home-Gateway to the end host.  The next three look
at that information for ways to send a message-in-a-bottle to the host
to enable mdns.

 c) Friendly ISP: magic DNS server IP address

    Robert Elz already covered most of the aspects of this.
    I'm not thinking of specifying a multicast address as the DNS
    server IP address, but something like 127.127.127.127 (say).

    Doesn't work if there is a DNS proxy in the home gateway
    (ie if the magic addr is stripped)

    Proposal: that mdns be enabled if magic ip address is one of the
	      DNS server addresses

 d) Friendly ISP: DNS search list

    We have done this one to death on the list.  Not much support.
    Loads root (or gTLD) nameservers with lookups for .local.arpa
    on clients which don't understand.

    Proposal: mdns enabled if .local.arpa in the DNS search list

 e) Friendly ISP: set the host domain name to local.arpa

    Mosts hosts don't seem to do a lot with their domain name setting.
    Setting the domain name to the ISP domain name seems pretty
    pointless for a home network behind a gateway.  The ISP can still
    use the DNS search list to configure clients.

    Proposal: mdns enabled when a DHCP response is received with a
	      domain name of .local.arpa.

    It is probably less likely that the domain name will make it
    through the home gateway than an IP address or search list.
    A gateway should take care not to enable mdns on the ISP side,
    perhaps a restriction to private addresses may be appropriate.

    I *think* this could even be per-interface, even though the domain
    name is probably per-host configuration.


My preferences are for:

  b) on the grounds that mdns will be enabled without ISPs or
     non-technical consumers needing to do anything and will remain
     disabled in most (many?) configured environments

  e) on the grounds that ISPs can do this without needing to update
     their software


Comments?  Better ideas?

regards
	aidan
____
:wq!


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 24 11:33:22 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA19254
	for <dnsext-archive@lists.ietf.org>; Mon, 24 Sep 2001 11:33:20 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15lX6U-000Gzo-00
	for namedroppers-data@psg.com; Mon, 24 Sep 2001 07:53:02 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15lX6U-000Gzi-00
	for namedroppers@ops.ietf.org; Mon, 24 Sep 2001 07:53:02 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15lX6T-0005vZ-00
	for namedroppers@ops.ietf.org; Mon, 24 Sep 2001 07:53:01 -0700
Message-Id: <200109241309.JAA03028@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce: ;
Cc: namedroppers@ops.ietf.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-dnsext-mdns-05.txt
Date: Mon, 24 Sep 2001 09:09:23 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the DNS Extensions Working Group of the IETF.

	Title		: Multicast DNS
	Author(s)	: L. Esibov, B. Aboba, D. Thaler
	Filename	: draft-ietf-dnsext-mdns-05.txt
	Pages		: 17
	Date		: 20-Sep-01
	
Today, with the rise of home networking, there are an increasing number
of ad-hoc networks operating without a DNS server. In order to allow DNS
name resolution in such environments, the use of a multicast DNS is
proposed.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-dnsext-mdns-05.txt

To remove yourself from the IETF Announcement list, send a message to 
ietf-announce-request with the word unsubscribe in the body of the message.

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-dnsext-mdns-05.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-dnsext-mdns-05.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20010920115820.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-dnsext-mdns-05.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-dnsext-mdns-05.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20010920115820.I-D@ietf.org>

--OtherAccess--

--NextPart--




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 24 12:26:50 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA21968
	for <dnsext-archive@lists.ietf.org>; Mon, 24 Sep 2001 12:26:48 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15lYBP-000Hz9-00
	for namedroppers-data@psg.com; Mon, 24 Sep 2001 09:02:11 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15lYBP-000Hz3-00
	for namedroppers@ops.ietf.org; Mon, 24 Sep 2001 09:02:11 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15lYBP-0009N6-00
	for namedroppers@ops.ietf.org; Mon, 24 Sep 2001 09:02:11 -0700
Date: Mon, 24 Sep 2001 08:44:03 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Aidan Williams <Aidan_Williams-A15677@email.mot.com>
cc: namedroppers@ops.ietf.org
Subject: Re: Enabling mDNS in the home environment
In-Reply-To: <3BAECDE8.7DA9E8C0@email.mot.com>
Message-ID: <Pine.BSF.4.21.0109240820530.22510-100000@internaut.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

>     (could be any rfc1918 prefix, but is typically 192.168/16)
> 

Actually, I've most often seen 10/8 used for this. 

>     I don't know of a home gateway with a DNS server supporting DDNS.

Actually there are a few that do, including Microsoft ICS. They handle
this via integration of the DHCP server with the DNS proxy. That way the
DNS proxy doesn't have to handle dynamic update. 

>   - DDNS to a DNS server in the gateway box

The gateway DNS server probably won't support dynamic update. About the
best it can do is answer queries for RRs dynamically updated via the
built-in DHCP server. 

>     Hosts could use DDNS directly, or the DHCP server in the gateway
>     box could do it.  A DHCP server could use a supplied hostname to
>     populate .local.arpa. some other way too.

I think it's unlikely that gateways will implement Dynamic Update. So the
DHCP server should be assumed to do it. Which brings up the
question: "What about IPv6, with stateless autoconfig and no DHCPv6? How
would DDNS work in this environment?"

>     Requires a home gateway vendor to implement a DNS server, and
>     integrate it with DHCP or DDNS.  Multiple DNS servers in the home
>     desiring to manage .local.arpa. are tricky.  How do they
>     co-operate?

Integrating the built-in DHCP server with the DNS server on the gateway so
as to provide DDNS isn't much work, and so to my mind this is what the
home gateways should do. Thus in some respects the goal is to enable mDNS
when the DNS & DHCP servers *don't* support dynamic DNS. 


>     DDNS isn't mandatory in draft-aboba-dhc-mini-03.txt, which is the
>     nearest thing I know of to a spec for this kind of home gateway.


It is a good idea to support DDNS, though making it mandatory might be too
strong. 

>         (I haven't found an ISP yet willing to do DDNS)

Neither have I.


>     I question whether most ISPs have any desire to be involved in the
>     management of names and addresses for every device in the home.

I doubt it. 

>     Proposal: that mdns be disabled when NetBIOS naming is configured,
> 	      enabled otherwise

Might there not be enterprise customers that would want to turn off *both*
NetBIOS and mDNS? In my experience, customers who are able to turn off
NetBIOS also tend to have sophisticated DNS and DHCP infrastructures,
including support for DDNS. That's why they don't feel that they need
NetBIOS -- dynamic naming is already being provided via DDNS. 

I would suspect that such customers also wouldn't need mDNS. 

>     Doesn't work if there is a DNS proxy in the home

Rather than "striping DNS servers" included in the DNS server option of
DHCP, it probably makes more sense to add them to the end of the
list. That is the kind of thing that probably should be in the mini-DHCP
spec.

> Comments?  Better ideas?
> 

Thanks for posting this. I think you've gone through most of the
scenarios. 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Mon Sep 24 17:37:50 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA01734
	for <dnsext-archive@lists.ietf.org>; Mon, 24 Sep 2001 17:37:48 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15lcrm-000Mjc-00
	for namedroppers-data@psg.com; Mon, 24 Sep 2001 14:02:14 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15lcrm-000MjW-00
	for namedroppers@ops.ietf.org; Mon, 24 Sep 2001 14:02:14 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15lcrl-0001oZ-00
	for namedroppers@ops.ietf.org; Mon, 24 Sep 2001 14:02:13 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Mon, 24 Sep 2001 13:28:20 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Erik Guttman <Erik.Guttman@Sun.COM>
cc: Aidan Williams <Aidan_Williams-A15677@email.mot.com>,
        namedroppers@ops.ietf.org, aboba@internaut.com
Subject: Re: New(er) mDNS configuration language
In-Reply-To: <Roam.SIMC.2.0.6.1001078898.12056.erikg@ehdb03-home.germany>
Message-ID: <Pine.BSF.4.21.0109241326480.22909-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

How is this updated text?

3.1.  mDNS configuration

Multicast DNS usage can be configured manually or automatically.  On
interfaces where no manual or automatic configuration has been
performed, multicast DNS is enabled by default.

For IPv6, the stateless DNS discovery mechanisms described in [19] can
be used to discover whether multicast DNS is globally enabled or
disabled.

Where DHCPv4 or DHCPv6 is implemented, DHCP options can be used to
configure multicast DNS on an interface.  The mDNS Enable Option,
described in [6], can be used to explicitly enable or disable use of
multicast DNS on an interface, as well as to specify the order in which
DNS and multicast DNS is used on that interface.

The mDNS Enable Option affects only DNS resolver behavior, that is, how
DNS resolution is performed, and whether multicast DNS is used.  The
mDNS Enable Option does not determine whether or in which order DNS
itself is used for name resolution.  This can be specified, for example,
using the Name Service Search Option for DHCP, RFC 2937 [3], which can
be used to globally determine where DNS is used within the name service
search order.

If an interface has been configured via any automatic configuration
mechanism which is able to supply DNS configuration information, then
multicast DNS MUST NOT be used on that interface unless it has been
explicitly enabled, whether via that mechanism or any other. This
ensures that upgraded hosts do not change their default behavior,
without requiring the source of the configuration information to be
simultaneously updated.  This implies that on the interface, the host
will neither listen on the DNS LINKLOCAL multicast address, nor will it
send queries to that address.  For a DNS server, automatic configuration
mechanisms MUST NOT  enable multicast DNS on any interface.

On Fri, 21 Sep 2001, Erik Guttman wrote:

> Bernard, Aidan,
> 
> > > However, if it is going to work this way, I think that the
> > > interactions between mdns-enable and the search order need to be
> > > spelled out (i.e. Erik's table probably need to be included
> > > somewhere), even if people don't bother to implement rfc2937.
> > > 
> > 
> > Erik - do you have some proposed text to insert?
> 
> I suggest:
> 
>   The mDNS Enable Option specifies the order in which DNS and multicast
>   DNS is done.  The mDNS Enable Option affects only DNS resolver behavior,
>   that is, how DNS resolution is performed, whether multicast DNS is used.  
> 
>   The mDNS Enable Option does not determine whether or in which order DNS 
>   itself is used for name resolution.  This can be specified, for example, 
>   using the Name Service Search Option for DHCP [RFC2937].
> 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 25 09:35:23 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA02753
	for <dnsext-archive@lists.ietf.org>; Tue, 25 Sep 2001 09:35:22 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15lrtV-000BfN-00
	for namedroppers-data@psg.com; Tue, 25 Sep 2001 06:05:01 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15lrtV-000BfH-00
	for namedroppers@ops.ietf.org; Tue, 25 Sep 2001 06:05:01 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15lrtU-000ORI-00
	for namedroppers@ops.ietf.org; Tue, 25 Sep 2001 06:05:00 -0700
Message-Id: <200109251100.HAA26958@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce: ;
Cc: namedroppers@ops.ietf.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-dnsext-gss-tsig-03.txt
Date: Tue, 25 Sep 2001 07:00:53 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the DNS Extensions Working Group of the IETF.

	Title		: GSS Algorithm for TSIG (GSS-TSIG)
	Author(s)	: S. Kwan, P. Garg, J. Gilroy, L. Esibov, 
                          J. Westhead, R. Hall
	Filename	: draft-ietf-dnsext-gss-tsig-03.txt
	Pages		: 21
	Date		: 24-Sep-01
	
The TSIG protocol provides transaction level authentication for DNS.
TSIG is extensible through the definition of new algorithms.  This
document specifies an algorithm based on the Generic Security Service
Application Program Interface (GSS-API) (RFC2743).

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-dnsext-gss-tsig-03.txt

To remove yourself from the IETF Announcement list, send a message to 
ietf-announce-request with the word unsubscribe in the body of the message.

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-dnsext-gss-tsig-03.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-dnsext-gss-tsig-03.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20010924124036.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-dnsext-gss-tsig-03.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-dnsext-gss-tsig-03.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20010924124036.I-D@ietf.org>

--OtherAccess--

--NextPart--




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Tue Sep 25 15:10:51 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA12337
	for <dnsext-archive@lists.ietf.org>; Tue, 25 Sep 2001 15:10:51 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15lxBm-000G8e-00
	for namedroppers-data@psg.com; Tue, 25 Sep 2001 11:44:14 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15lxBm-000G8Y-00
	for namedroppers@ops.ietf.org; Tue, 25 Sep 2001 11:44:14 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15lxBl-000FX0-00
	for namedroppers@ops.ietf.org; Tue, 25 Sep 2001 11:44:13 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Tue, 25 Sep 2001 11:30:41 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: namedroppers@ops.ietf.org
Subject: [mDNS] Removal of ".local.arpa" namespace?
Message-ID: <Pine.BSF.4.21.0109251128070.24493-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

It has been pointed out that now that we are no longer using the domain
searchlist to configure mDNS, that continuing to use the
".local.arpa" namespace doesn't make sense. 

Assuming that you could use mDNS to query for *any* name, is there a
reason for continuing to include ".local.arpa" in the draft? 



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 01:42:28 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA26758
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 01:42:28 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15m6zw-000Nyx-00
	for namedroppers-data@psg.com; Tue, 25 Sep 2001 22:12:40 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15m6zv-000Nyr-00
	for namedroppers@ops.ietf.org; Tue, 25 Sep 2001 22:12:39 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15m6zv-000NEQ-00
	for namedroppers@ops.ietf.org; Tue, 25 Sep 2001 22:12:39 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Randy Bush <randy@psg.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
Message-Id: <E15m6zI-000NCA-00@rip.psg.com>
Date: Tue, 25 Sep 2001 22:12:00 -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

two week last call starts for draft-ietf-dnsext-unknown-rrs-01.txt for
proposed standard.

randy


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 04:44:51 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA10865
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 04:44:51 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mA03-0000XR-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 01:24:59 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mA02-0000XL-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 01:24:58 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mA02-0006jd-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 01:24:58 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 08:44:58 +0200 (MET DST)
From: Erik Guttman <Erik.Guttman@Sun.COM>
Subject: Re: [mDNS] Removal of ".local.arpa" namespace?
To: Bernard Aboba <aboba@internaut.com>
Cc: namedroppers@ops.ietf.org
In-Reply-To: "Your message with ID" <Pine.BSF.4.21.0109251128070.24493-100000@internaut.com>
Message-ID: <Roam.SIMC.2.0.6.1001486698.4331.erikg@ehdb03-home.germany>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> It has been pointed out that now that we are no longer using the domain
> searchlist to configure mDNS, that continuing to use the
> ".local.arpa" namespace doesn't make sense. 
> 
> Assuming that you could use mDNS to query for *any* name, is there a
> reason for continuing to include ".local.arpa" in the draft? 

There are very good reasons not to include it:  There would have
to be special handling of this arpa domain (by the arpa TLD servers)
which we have not come to any resolution about.  Further, there will
still be a prolonged discussion about whether this particular use
of the arpa TLD is appropriate.  

If there is a desire to create an administratively local namespace
for specific use (say, only link-local, or link-local and admin
scope) this should be defined in a separate document.  

Erik



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 09:01:50 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA16790
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 09:01:49 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mDzE-0004AQ-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 05:40:24 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mDzD-0004AK-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 05:40:23 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mDzD-000J8G-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 05:40:23 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 11:17:45 +0200
From: bert hubert <ahu@ds9a.nl>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
Message-ID: <20010926111745.A10616@outpost.ds9a.nl>
References: <E15m6zI-000NCA-00@rip.psg.com>
In-Reply-To: <E15m6zI-000NCA-00@rip.psg.com>; from randy@psg.com on Tue, Sep 25, 2001 at 10:12:00PM -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Tue, Sep 25, 2001 at 10:12:00PM -0700, Randy Bush wrote:
> two week last call starts for draft-ietf-dnsext-unknown-rrs-01.txt for
> proposed standard.

Section 8 mentions addidtional processing. I'm afraid that this is a very
real problem - a lot records of will want to have additional processing,
which can't be expected of a server not understanding the record.

Furthermore, compression is a problem. An AXFR in 'big packet' format will
have label compression. The proposal to forbid compression of RRs which are
not well-known would prevent the more interesting applications which
require compression to reasonably fit in 512 octet packets.

We will implement this in our nameserver, even if only for testing purposes
- it allows us to serve records for which we have no support yet. As we have
no ascii zone representation, but store data directly in a relational
database, section 5 changes slightly of course, but not relevantly so.

But this proposal appears to be a slightly crippled proposal. It rules out:

	non-well-known RRtypes which need additional processing.
	non-well-known RRtypes which need compression 

The problem may not have an easy solution I'm afraid.

Regards,

bert hubert
PowerDNS

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 09:01:52 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA16801
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 09:01:51 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mDy2-00049J-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 05:39:10 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mDy1-00049D-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 05:39:09 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mDy1-000J4e-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 05:39:09 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: Erik Guttman <Erik.Guttman@Sun.COM>
Cc: Bernard Aboba <aboba@internaut.com>, namedroppers@ops.ietf.org
In-reply-to: Erik.Guttman's message of Wed, 26 Sep 2001 08:44:58 +0200.
      <Roam.SIMC.2.0.6.1001486698.4331.erikg@ehdb03-home.germany> 
Subject: Re: [mDNS] Removal of ".local.arpa" namespace? 
From: itojun@iijlab.net
Date: Wed, 26 Sep 2001 17:58:24 +0900
Message-ID: <8024.1001494704@itojun.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>> It has been pointed out that now that we are no longer using the domain
>> searchlist to configure mDNS, that continuing to use the
>> ".local.arpa" namespace doesn't make sense. 
>> Assuming that you could use mDNS to query for *any* name, is there a
>> reason for continuing to include ".local.arpa" in the draft? 
>
>There are very good reasons not to include it:  There would have
>to be special handling of this arpa domain (by the arpa TLD servers)
>which we have not come to any resolution about.  Further, there will
>still be a prolonged discussion about whether this particular use
>of the arpa TLD is appropriate.  
>
>If there is a desire to create an administratively local namespace
>for specific use (say, only link-local, or link-local and admin
>scope) this should be defined in a separate document.  

	I believe it is still useful to use ".local.arpa" for mDNS.  It would
	be a disaster for various mDNS-enabled node to claim authority for
	"foo.itojun.org".  I don't care if it is defined in a separate draft,
	or in a single draft, but I would like to see it kept in some way,
	to protect normal DNS namespace from getting contaminated with mDNS
	results.

itojun


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 09:01:54 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA16804
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 09:01:52 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mE4V-0004G5-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 05:45:51 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mE4V-0004Fz-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 05:45:51 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mE4V-000JPC-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 05:45:51 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 12:51:59 +0200 (MET DST)
From: Erik Guttman <Erik.Guttman@Sun.COM>
Subject: Re: [mDNS] Removal of ".local.arpa" namespace? 
To: itojun@iijlab.net
Cc: Erik Guttman <Erik.Guttman@Sun.COM>, Bernard Aboba <aboba@internaut.com>,
        namedroppers@ops.ietf.org
In-Reply-To: "Your message with ID" <8024.1001494704@itojun.org>
Message-ID: <Roam.SIMC.2.0.6.1001501519.32499.erikg@ehdb03-home.germany>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> 	I believe it is still useful to use ".local.arpa" for mDNS.  It would
> 	be a disaster for various mDNS-enabled node to claim authority for
> 	"foo.itojun.org".  I don't care if it is defined in a separate draft,
> 	or in a single draft, but I would like to see it kept in some way,
> 	to protect normal DNS namespace from getting contaminated with mDNS
> 	results.

Itojun,

This active attack threat (masquerading) is indeed a concern.
Thank you for bringing it up again for discussion.  

I submit that this will be present regardless of whether mdns,
and whether '.local.arpa' is used as currently defined, or not.

Consider that on a LAN, any host can respond to DNS requests
as if they were the DNS server by evesdropping and generating
the appropriate IP packet reply.  DNSsec or IPsec or some other
measure would be needed to prevent this.  This threat is independent 
of whether multicast DNS is used, or which namespace.

A larger multicast scope would ease an attacker's job - as he/she
could be anywhere within the extent of multicast routing rather
than on the forwarding path of the DNS request.  We are
not currently discussing mDNS using larger than link-local scope
and even so, this is only worsens the problem a bit by degree but
not by kind.

Also, using .local.arpa (as discussed by previous drafts of
draft-ietf-dnsext-mdns) does not reduce the risk of masquerade
attacks or even unwanted aliasing behavior.  'foo.itojun.org' would
be resolved as 'foo.itojun.org.local.arpa', and the mdns server on 
the host with authority for 'foo.itojun.org' would respond to
'foo.itojun.org.local.arpa'.  This aliasing buys one nothing.

Last, any host besides the one you assigned the name 'foo.itojun.org'
responding to that name would be doing so in violation of the
protocol which prevents duplicate name assignment.

This brings up an interesting questions I think:

Shouldn't a host H1 with an administratively assigned name defend
it against another host H2 which is using the same name, even if
H1 discovers this fact after H2 has already claimed the name?
I distinguish between two kinds of names - those assigned and those 
'invented.' In this case shouldn't the latter category should be 
prepared to relinquish the name, the former not?   This is not an
arbitrary nor philosophical distinction (assigned names vs. zeroconf 
generated names).

Regards,

Erik



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 10:43:03 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA20315
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 10:43:02 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mFTn-0006N0-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 07:16:03 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mFTn-0006Mu-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 07:16:03 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mFTn-000NtV-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 07:16:03 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 10:12:06 -0400 (EDT)
From: Olafur Gudmundsson <ogud@ogud.com>
To: bert hubert <ahu@ds9a.nl>
cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010926111745.A10616@outpost.ds9a.nl>
Message-ID: <Pine.BSF.4.21.0109260923260.71704-100000@hlid.dc.ogud.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Wed, 26 Sep 2001, bert hubert wrote:

> On Tue, Sep 25, 2001 at 10:12:00PM -0700, Randy Bush wrote:
> > two week last call starts for draft-ietf-dnsext-unknown-rrs-01.txt for
> > proposed standard.
> 
> Section 8 mentions addidtional processing. I'm afraid that this is a very
> real problem - a lot records of will want to have additional processing,
> which can't be expected of a server not understanding the record.

Additional section processing allways has been optional, all this document 
is to point out that it is also dependant on modified server deployment. 

> 
> Furthermore, compression is a problem. An AXFR in 'big packet' format will
> have label compression. The proposal to forbid compression of RRs which are
> not well-known would prevent the more interesting applications which
> require compression to reasonably fit in 512 octet packets.

This only forbids compression in the right hand side (rdata) of new
rr types. The left hand side can still be compressed. 
If you need large packets use ENDS to increase the packet size, EDNS
is basicly mandatory for any implementation that supports DNSSEC and
A6 (see message-size document that is pending RFC publication). 

> 
> We will implement this in our nameserver, even if only for testing purposes
> - it allows us to serve records for which we have no support yet. As we have
> no ascii zone representation, but store data directly in a relational
> database, section 5 changes slightly of course, but not relevantly so.

Presentation format is optional and is mainly aimed at tools used to 
debug and test DNS setups.

> 
> But this proposal appears to be a slightly crippled proposal. It rules out:
> 
> 	non-well-known RRtypes which need additional processing.
> 	non-well-known RRtypes which need compression 

What this document summarizes is what has been learned over the last
15 years in what area of compression (bad idea to allow it in rdata), 
and point out that additional section processing requires server upgrades. 

Unless someone writes a server that reads from a configuration file 
format of RR types and the additional section rules for that type. 
Now supporting a new type is a simple patch to this file. 

Compression of rdata is a double eged sword, it reduces the  ize of
answers but larger packtes render this less important. Compression
increases work and complexity of servers and resolvers making the system 
less reliable.

> The problem may not have an easy solution I'm afraid.

This document specifies the best possible solution to this problem,
unless we of course ban new types :-) 

	Olafur



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 10:44:57 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA20348
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 10:44:55 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mFck-0006Yo-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 07:25:18 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mFck-0006Ye-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 07:25:18 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mFck-000OLp-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 07:25:18 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 07:11:26 -0700 (PDT)
From: Bernard Aboba <aboba@internaut.com>
To: Erik Guttman <Erik.Guttman@Sun.COM>
cc: itojun@iijlab.net, namedroppers@ops.ietf.org
Subject: Re: [mDNS] Removal of ".local.arpa" namespace? 
In-Reply-To: <Roam.SIMC.2.0.6.1001501519.32499.erikg@ehdb03-home.germany>
Message-ID: <Pine.BSF.4.21.0109260700460.25790-100000@internaut.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> > 	I believe it is still useful to use ".local.arpa" for mDNS.  It would
> > 	be a disaster for various mDNS-enabled node to claim authority for
> > 	"foo.itojun.org".  I don't care if it is defined in a separate draft,
> > 	or in a single draft, but I would like to see it kept in some way,
> > 	to protect normal DNS namespace from getting contaminated with mDNS
> > 	results.

The main thing that ".local.arpa" provides is
the ability of two hosts with different searchlists to resolve the name
non-fully qualified domain names, such as "foo". 

> This active attack threat (masquerading) is indeed a concern.
> Thank you for bringing it up again for discussion.  
> 
> I submit that this will be present regardless of whether mdns,
> and whether '.local.arpa' is used as currently defined, or not.
> 

Indeed, the current draft doesn't prohibit an mDNS-enabled node from
answering a query for foo.itojun.org, foo.itojun.org.local.arpa, etc.

> Last, any host besides the one you assigned the name 'foo.itojun.org'
> responding to that name would be doing so in violation of the
> protocol which prevents duplicate name assignment.

Actually what is prohibited is not duplicate name assignment, but
duplicate name usage. 

> Shouldn't a host H1 with an administratively assigned name defend
> it against another host H2 which is using the same name, even if
> H1 discovers this fact after H2 has already claimed the name?

This occurred to me today as I encountered a name conflict on our
corporate network. Apparently lots of people like to name their computers
based on the manufacturer and model # (e.g. dell8100). I had registered my
machine, but someone else also picked the name, and I had to change mine
(not a trivial procedure, by the way, since I had to obtain new machine
certs for use with IPsec, IEEE 802.1X, etc.). Took me about 15 minutes to
complete the name change and make sure everything was working again. 

> I distinguish between two kinds of names - those assigned and those 
> 'invented.' In this case shouldn't the latter category should be 
> prepared to relinquish the name, the former not?   This is not an
> arbitrary nor philosophical distinction (assigned names vs. zeroconf 
> generated names).

Indeed.



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 10:48:17 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA20429
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 10:48:15 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mFgm-0006d2-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 07:29:28 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mFgl-0006cv-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 07:29:27 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mFgl-000OYl-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 07:29:27 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 16:27:30 +0200
From: bert hubert <ahu@ds9a.nl>
To: Olafur Gudmundsson <ogud@ogud.com>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
Message-ID: <20010926162730.B15728@outpost.ds9a.nl>
References: <20010926111745.A10616@outpost.ds9a.nl> <Pine.BSF.4.21.0109260923260.71704-100000@hlid.dc.ogud.com>
In-Reply-To: <Pine.BSF.4.21.0109260923260.71704-100000@hlid.dc.ogud.com>; from ogud@ogud.com on Wed, Sep 26, 2001 at 10:12:06AM -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Wed, Sep 26, 2001 at 10:12:06AM -0400, Olafur Gudmundsson wrote:

> > Section 8 mentions addidtional processing. I'm afraid that this is a very
> > real problem - a lot records of will want to have additional processing,
> > which can't be expected of a server not understanding the record.
> 
> Additional section processing allways has been optional, all this document 
> is to point out that it is also dependant on modified server deployment. 

I've never seen additional processing described as optional, although I may
not have looked all that well. I do know that most implementations are
resilient in the face of nameservers that do not perform additional
processing. It does cause a lot of additional traffic though.

> > The problem may not have an easy solution I'm afraid.
> 
> This document specifies the best possible solution to this problem,
> unless we of course ban new types :-) 

Oh I think it is a very good proposal - it may be the best possible. It does
not solve the problem very well however. Rephrasing the limitations you
mentioned, this proposal:

	- Suggests that new RRtypes do not use compression on the right hand
	  side

	- Suggests that everybody interested in new RRtypes use an EDNS
	  compliant nameserver

Regards,

bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
Netherlabs BV / Rent-a-Nerd.nl           - Nerd Available -
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 15:53:51 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA27429
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 15:53:49 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mJtm-000COD-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 11:59:10 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mJtm-000CO7-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 11:59:10 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mJtm-000BvW-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 11:59:10 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Randy Bush <randy@psg.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: wg last call for draft-ietf-dnsext-dhcid-rr-03.txt to ps
Message-Id: <E15mJrH-000Bmg-00@rip.psg.com>
Date: Wed, 26 Sep 2001 11:56:35 -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

this starts a two week wg last call on draft-ietf-dnsext-dhcid-rr-03.txt
for proposed standard.

randy


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 16:02:02 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA27565
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 16:02:01 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mK6i-000Ccr-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 12:12:32 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mK6i-000Cck-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 12:12:32 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mK6i-000Cbj-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 12:12:32 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 12:04:36 -0700 (PDT)
Message-Id: <200109261904.MAA03755@gulag.araneus.fi>
To: bert hubert <ahu@ds9a.nl>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010926111745.A10616@outpost.ds9a.nl>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926111745.A10616@outpost.ds9a.nl>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

bert hubert writes:
> Section 8 mentions addidtional processing. I'm afraid that this is a very
> real problem - a lot records of will want to have additional processing,
> which can't be expected of a server not understanding the record.

As you said, there is no easy solution.  Servers can't do
type-specific additional section processing for record types that are
not defined yet.  We only have the choice between rejecting the new
records completely, like many existing servers do, or serving them
without doing additional section processing.  I strongly believe the
latter is more useful.

Servers written after a new RR type has been defined can and will do
additional section processing for them.  The text in section 8 is to
remind those defining new RR types that no matter what they specify,
existing servers in conformance with the draft will transmit the RRs
without doing additional section processing, in preference to dropping
them completely.

> Furthermore, compression is a problem. An AXFR in 'big packet' format will
> have label compression. The proposal to forbid compression of RRs which are
> not well-known would prevent the more interesting applications which
> require compression to reasonably fit in 512 octet packets.

Servers also can't do type-specific compression of records types that
are not defined yet.  If we allow compression in new types, it is
simply impossible for existing servers to transmit them correctly - if
they are not dropped, they will be silently corrupted.  When deploying
a new RR type, a minor increase in packet size due to the lack of name
compression is a small price to pay for not having to upgrade every
caching DNS server in the world.

I also think the benefits of RDATA compression in new types are
overrated.  A large fraction of the compression gain comes from
compressing owner names (the left hand side of the RR) rather than
names embeded in the RDATA, and owner names will continue to be
compressed even in the case of new RR types.  Also, it seems likely
that most new RR types will contain application payloads consisting
mostly of data other than domain names, and therefore would not
benefit significantly from compression.

At the Oslo DNSIND meeting, Randy Bush asked the room whether anyone
had any pressing need for any kind of compression inside the RR data
fields of anything but the NS records of the root zone.  No one did.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 16:04:11 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA27621
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 16:04:10 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mK8m-000Cgw-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 12:14:40 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mK8m-000Cgp-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 12:14:40 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mK8l-000Cik-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 12:14:39 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 12:40:54 -0400
From: "Nicolae, Dan" <Dan.Nicolae@usa.xerox.com>
Subject: RE: [mDNS] Removal of ".local.arpa" namespace?
To: "'Erik Guttman'" <Erik.Guttman@Sun.COM>, itojun@iijlab.net
Cc: Bernard Aboba <aboba@internaut.com>, namedroppers@ops.ietf.org
Message-id: <3820E1EF15CCD411B4E600508BAED1F45BCA4D@usa0111ms2.eng.mc.xerox.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> It has been pointed out that now that we are no longer using 
> the domain
> searchlist to configure mDNS, that continuing to use the
> ".local.arpa" namespace doesn't make sense

> 	I believe it is still useful to use ".local.arpa" for 
> mDNS.  It would
> 	be a disaster for various mDNS-enabled node to claim 
> authority for
> 	"foo.itojun.org".  

> 
> This active attack threat (masquerading) is indeed a concern.
> Thank you for bringing it up again for discussion.  
> 

Is mDNS ever going to coexist with standard DNS on the same client? Is a DNS
server ever going to respond to multicast queries? Is mDNS ever going to go
beyond linklocal? 

If yes, then this is not only a masquerading issue. This namespace
contamination, authority conflict, configured vs. zero-configured names, or
however we may call it, *may* become an administrative nightmare in the
future. Even when troubleshooting NetBIOS name resolution is useful to
distinguish between names resolved through broadcasts vs. names resolved
through name server (i.e. nbtstat -r).

On the other hand, if the answer to the questions above is no, then using
the NS/SOA records, DDNS and maybe even multicast to support mDNS is
overkill. So I assume that at least some people hope that the answer those
questions is yes. 

However, we may assume that an intelligent proxy would fill the gap between
mDNS and DNS and take care of any potential coexistence conflicts:
mdns-05: "...Propagation of multicast DNS packets within the local subnet is
considered sufficient to enable DNS name resolution in small adhoc
networks. The assumption is that if a network has a router, then the
network either has a DNS server or the router can function as a DNS
proxy, possibly implementing dynamic DNS..."

The fact that we restrict for now mDNS to linklocal does not mean that at
some point the mDNS traffic won't escape to the Internet, even if it is
repackaged as legitimate unicast traffic.

What if the network has that smart router that accepts DDNS registrations
from the linklocal hosts and also resolves names for hosts outside of
linklocal? A local host manually configured as "foo.itojun.org" would
register its name with the router and the real "foo.itojun.org" on the
Internet would never be found.

Also, are users expected to name their PC using a FQDN? I think most of them
would simply name it "mypc" and not worry about the rest. What is going to
happen then? The router would have to accept registrations for *any* FQDN
and become authoritative for *any* zone, including the root and top level
domains. How about the zillion of smart devices that will most likely do
mDNS and ship with a default name? What would be a good default domain? What
if a home user has the idea of naming his new printer with a name under the
printer's manufacturer's DNS domain?

Etc. It is hard to foresee all potential problems, but I *feel* that there
is something wrong with completely giving up control over the DNS authority.
There must be way to distinguish between administratively configured and the
selfconfigured/user configured data. 

"local.arpa" was a quick and dirty candidate for meeting this goal. Although
I personally don't like it, we can't afford to give up on it without a
replacement solution. A while ago I suggested using a different class to
flag the non-administratively configured DNS data. The need of a better
authority discovery mechanism was also pointed out. But, the way mDNS looks
today, and if it is to be extended tomorrow, I think we're stuck with
'local.arpa'


dan






to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 17:13:08 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA28850
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 17:13:07 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mLfM-000Esr-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 13:52:24 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mLfL-000Esl-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 13:52:23 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mLfL-000Hgx-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 13:52:23 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 22:47:05 +0200
From: bert hubert <ahu@ds9a.nl>
To: Andreas Gustafsson <gson@isc.org>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
Message-ID: <20010926224705.A19138@outpost.ds9a.nl>
References: <E15m6zI-000NCA-00@rip.psg.com> <20010926111745.A10616@outpost.ds9a.nl> <200109261904.MAA03755@gulag.araneus.fi>
In-Reply-To: <200109261904.MAA03755@gulag.araneus.fi>; from gson@isc.org on Wed, Sep 26, 2001 at 12:04:36PM -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Wed, Sep 26, 2001 at 12:04:36PM -0700, Andreas Gustafsson wrote:

> a new RR type, a minor increase in packet size due to the lack of name
> compression is a small price to pay for not having to upgrade every
> caching DNS server in the world.

Well, a caching DNS server doesn't need the RDATA type - I think this
proposal has the most merit when doing AXFRs to older slave servers.

Just a thought, It may be useful to state that authoritative nameservers
SHOULD warn against RDATA records which it actually does understand
natively - I see possibilities for havoc otherwise.

> At the Oslo DNSIND meeting, Randy Bush asked the room whether anyone
> had any pressing need for any kind of compression inside the RR data
> fields of anything but the NS records of the root zone.  No one did.

RP comes to mind, but as nobody is using that :-) RDATA may be more useful
than I originally thought. 

Regards,

bert hubert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
Netherlabs BV / Rent-a-Nerd.nl           - Nerd Available -
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 17:15:44 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA28892
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 17:15:42 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mLcM-000Emu-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 13:49:18 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mLcL-000Emo-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 13:49:17 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mLcL-000HXG-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 13:49:17 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BB23A32.4FB62E0D@daimlerchrysler.com>
Date: Wed, 26 Sep 2001 16:27:30 -0400
From: Kevin Darcy <kcd@daimlerchrysler.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <E15m6zI-000NCA-00@rip.psg.com>
		<20010926111745.A10616@outpost.ds9a.nl> <200109261904.MAA03755@gulag.araneus.fi>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

gson@isc.org wrote:

> bert hubert writes:
> > Section 8 mentions addidtional processing. I'm afraid that this is a very
> > real problem - a lot records of will want to have additional processing,
> > which can't be expected of a server not understanding the record.
>
> As you said, there is no easy solution.  Servers can't do
> type-specific additional section processing for record types that are
> not defined yet.  We only have the choice between rejecting the new
> records completely, like many existing servers do, or serving them
> without doing additional section processing.  I strongly believe the
> latter is more useful.
>
> Servers written after a new RR type has been defined can and will do
> additional section processing for them.  The text in section 8 is to
> remind those defining new RR types that no matter what they specify,
> existing servers in conformance with the draft will transmit the RRs
> without doing additional section processing, in preference to dropping
> them completely.
>
> > Furthermore, compression is a problem. An AXFR in 'big packet' format will
> > have label compression. The proposal to forbid compression of RRs which are
> > not well-known would prevent the more interesting applications which
> > require compression to reasonably fit in 512 octet packets.
>
> Servers also can't do type-specific compression of records types that
> are not defined yet.  If we allow compression in new types, it is
> simply impossible for existing servers to transmit them correctly - if
> they are not dropped, they will be silently corrupted.  When deploying
> a new RR type, a minor increase in packet size due to the lack of name
> compression is a small price to pay for not having to upgrade everycaching DNS
> server in the world.

As a general rule, it's probably best to avoid name compression in new types, I
concur. But let's not be overly rigid about this: name compression should be
permissible where the client and server have already established a common
understanding of the new type through EDNS or some similar mechanism.


- Kevin




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 17:39:20 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA29491
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 17:39:19 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mM3R-000FRH-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 14:17:17 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mM3Q-000FRB-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 14:17:16 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mM3Q-000Ix2-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 14:17:16 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 14:15:53 -0700 (PDT)
Message-Id: <200109262115.OAA03851@gulag.araneus.fi>
To: bert hubert <ahu@ds9a.nl>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010926224705.A19138@outpost.ds9a.nl>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926111745.A10616@outpost.ds9a.nl>
	<200109261904.MAA03755@gulag.araneus.fi>
	<20010926224705.A19138@outpost.ds9a.nl>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> > a new RR type, a minor increase in packet size due to the lack of name
> > compression is a small price to pay for not having to upgrade every
> > caching DNS server in the world.
> 
> Well, a caching DNS server doesn't need the RDATA type - I think this
> proposal has the most merit when doing AXFRs to older slave servers.

What do you mean by "a caching DNS server doesn't need the RDATA
type"?  A caching server does need to know the specific layout of
every RR type that can be compressed or that requires specific
additional section processing.  Making it safe for caching servers to
transparently receive, cache and retransmit RRs of unknown type is the
primary motivation for the draft.

> Just a thought, It may be useful to state that authoritative nameservers
> SHOULD warn against RDATA records which it actually does understand
> natively - I see possibilities for havoc otherwise.

Such as?
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 17:59:34 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA29939
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 17:59:32 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mMVf-000G5L-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 14:46:27 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mMVf-000G5F-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 14:46:27 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mMVf-000KPl-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 14:46:27 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 14:42:39 -0700
From: Mark Gritter <mgritter@CS.Stanford.EDU>
To: namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
Message-ID: <20010926144239.A22605@Xenon.Stanford.EDU>
References: <E15m6zI-000NCA-00@rip.psg.com>
In-Reply-To: <E15m6zI-000NCA-00@rip.psg.com>; from randy@psg.com on Tue, Sep 25, 2001 at 10:12:00PM -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

I think this draft could use more in the way of requirements on caches
and forwarders--- "transparently" is not very precise.  Specifically, I think 
there is some need to look at requests and responses as whole units.

The language I would like to see is something like:

  A name server performing a recursive query must include any unknown 
  additional RRs which appear in the query.  The name server must not drop
  any unknown RRs appearing in the response it receives.

Motivation:

Extensions to DNS may involve more than new "data" types, as experienced with
DNSSEC.  Name servers being used for recursive queries do not need to 
understand the new RR types introduced as end-to-end protocol extensions,
but could interfere by dropping them from responses.  These RRs may not
be cacheable (which is an issue I'm not sure how to address in the standard---
I would be perfectly happy with not caching any response which includes
unknown RRs in the additional section.)

Merely bypassing such servers is not feasible in firewalled environments.

Most of the extensions I can think of that require such treatment would be 
controversial at best, I admit.
  -- Negotiation with load-balancing DNS.  One of the faults of such
     schemes is that they provide no way of requesting a different server
     if the one returned does not work out.  An RR in the additional section
     of a request could specify the address of a failed server, causing
     the load-balancer to return a different address.
  -- Encrypted responses.  Client public key could be included in request.
  -- Connection setup overlaid with name lookup.  (Yes, my research group
     has actually implemented this.  ;)

Without trying to argue the benefits of these individual examples, I'd 
argue that, realistically, there isn't an alternative to DNS, and so
we need to leave the door open to such protocol extensions that require
not separating unknown RRs from the known RRs they accompany.

Mark

-- 
Mark Gritter <mgritter@cs.stanford.edu>
http://www-cs-students.stanford.edu/~mgritter/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Wed Sep 26 18:00:24 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA29979
	for <dnsext-archive@lists.ietf.org>; Wed, 26 Sep 2001 18:00:23 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mMYE-000G8z-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 14:49:06 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mMYD-000G8t-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 14:49:05 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mMYD-000KYE-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 14:49:05 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 14:47:04 -0700 (PDT)
Message-Id: <200109262147.OAA03864@gulag.araneus.fi>
To: Kevin Darcy <kcd@daimlerchrysler.com>
CC: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <3BB23A32.4FB62E0D@daimlerchrysler.com>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926111745.A10616@outpost.ds9a.nl>
	<200109261904.MAA03755@gulag.araneus.fi>
	<3BB23A32.4FB62E0D@daimlerchrysler.com>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Kevin Darcy writes:
> As a general rule, it's probably best to avoid name compression in new
> types, I concur. But let's not be overly rigid about this: name
> compression should be permissible where the client and server have already
> established a common understanding of the new type through EDNS or some
> similar mechanism.

I doubt the benefits of additional compression will outweigh the cost
of such a mechanism.  But if the working group does decide to
standardize a mechanism for communicating extended compression support
from client to server, its specification can serve as an update to the
unknown-rrs specification, modifying the rules for where compression
is allowed.  There is no need to change the text of the unknown-rrs
draft itself in anticipation of this.
--
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 00:14:53 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA05810
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 00:14:53 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mSCY-000NZK-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 20:51:06 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mSCX-000NZD-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 20:51:05 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mSCX-000CKn-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 20:51:05 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BB251E7.E77BE8AB@daimlerchrysler.com>
Date: Wed, 26 Sep 2001 18:08:39 -0400
From: Kevin Darcy <kcd@daimlerchrysler.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <E15m6zI-000NCA-00@rip.psg.com>
		<20010926111745.A10616@outpost.ds9a.nl>
		<200109261904.MAA03755@gulag.araneus.fi>
		<3BB23A32.4FB62E0D@daimlerchrysler.com> <200109262147.OAA03864@gulag.araneus.fi>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

gson@isc.org wrote:

> Kevin Darcy writes:
> > As a general rule, it's probably best to avoid name compression in new types, I
> > concur. But let's not be overly rigid about this: name compression should be
> > permissible where the client and server have already established a common
> > understanding of the new type through EDNS or some similar mechanism.
>
> I doubt the benefits of additional compression will outweigh the cost
> of such a mechanism.  But if the working group does decide to
> standardize a mechanism for communicating extended compression support
> from client to server, its specification can serve as an update to the
> unknown-rrs specification, modifying the rules for where compression
> is allowed.  There is no need to change the text of the unknown-rrs
> draft itself in anticipation of this.

Sorry, I didn't make myself very clear. I wasn't referring to a generic "extended
compression support" negotiation feature, I simply meant that if a new record type
is only going to be used between clients and servers who are already known to
understand the type, e.g. the server only sends records of that type if the client
sent an EDNS0 OPTION requesting their inclusion, which the server understands and
with which it complies, then I see no reason why they can't use name compression in
their RDATA. Such "negotiated" compression is forbidden under the current draft
wording, and I find that too rigid.


- Kevin





to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 00:15:49 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id AAA05822
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 00:15:49 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mSEm-000Nbi-00
	for namedroppers-data@psg.com; Wed, 26 Sep 2001 20:53:24 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mSEm-000Nbc-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 20:53:24 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mSEm-000CRp-00
	for namedroppers@ops.ietf.org; Wed, 26 Sep 2001 20:53:24 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Wed, 26 Sep 2001 15:26:13 -0700 (PDT)
Message-Id: <200109262226.PAA03883@gulag.araneus.fi>
To: Mark Gritter <mgritter@CS.Stanford.EDU>
Cc: namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010926144239.A22605@Xenon.Stanford.EDU>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926144239.A22605@Xenon.Stanford.EDU>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Mark Gritter writes:
> I think this draft could use more in the way of requirements on caches
> and forwarders--- "transparently" is not very precise.  Specifically, I think 
> there is some need to look at requests and responses as whole units.
> 
> The language I would like to see is something like:
> 
>   A name server performing a recursive query must include any unknown 
>   additional RRs which appear in the query.  The name server must not drop
>   any unknown RRs appearing in the response it receives.

This is just not how caching servers work.  Caching servers don't
cache entire responses, they cache individual resource record sets.
What you are proposing is not a change to the unknown-rrs draft but
a fundamental change to how the DNS works.

> Name servers being used for recursive queries do not need to 
> understand the new RR types introduced as end-to-end protocol extensions,
> but could interfere by dropping them from responses.

There is no such thing as an end-to-end protocol extension in the DNS.
All existing extension that rely on special record types (EDNS0, TSIG,
SIG0) are hop-by-hop, i.e., they don't traverse caching servers.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 07:29:41 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA26536
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 07:29:40 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mYvq-00062U-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 04:02:18 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mYvp-00062N-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 04:02:17 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15mYvp-0007EO-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 04:02:17 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: 27 Sep 2001 05:23:08 -0000
Message-ID: <20010927052308.691.qmail@cr.yp.to>
From: "D. J. Bernstein" <djb@cr.yp.to>
To: namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <E15m6zI-000NCA-00@rip.psg.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Gustafsson's document says that caches ``SHOULD also decompress RRs of
type RP, AFSDB, RT, SIG, PX, NXT, NAPTR, and SRV.''

I objected to that text in December 2000, and I continue to object to
it, for several reasons:

   * RFC 1035 clearly prohibits compression except in owner names, NS
     data, CNAME data, PTR data, MX data, and SOA data. Compression in
     RP, AFSDB, etc. is inexcusable. There are no legitimate situations
     where decompression of RP, AFSDB, etc. will accomplish anything.

   * My cache, which is deployed at tens of thousands of sites, handles
     exactly the forms of compression allowed by RFC 1035. It does _not_
     decompress RP, AFSDB, etc. It works just fine.

   * Gustafsson's text violates RFC 2119, section 6. It is not necessary
     for interoperability. It is an abuse of the standards process.

   * BIND's support for bogus compressed records was exactly what led to
     the http://www.cert.org/advisories/CA-2000-20.html disaster. We do
     _not_ want more of these disasters.

The text should be replaced by the opposite recommendation. Implementors
should be encouraged to make their decompression code as limited as
possible, handling RFC 1035 compression and nothing else.

Gustafsson's only response to these points has been ``it's a SHOULD, not
a MUST,'' completely ignoring the substance of my objections.

I also object to this document's failure to point out that transparent
handling of new types is already required by the existing DNS standards.
See http://cr.yp.to/djbdns/newtypes.html.

---Dan


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 11:01:36 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA03227
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 11:01:32 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mcKM-000A0L-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 07:39:50 -0700
Received: from [147.28.4.2] (helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mcKL-000A0F-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 07:39:49 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mcKI-0009Az-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 07:39:46 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 16:14:22 +0200 (CEST)
From: Roy Arends <Roy.Arends@nominum.com>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010927052308.691.qmail@cr.yp.to>
Message-ID: <Pine.BSF.4.33.0109271507140.12760-100000@node10c4d.a2000.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

                There I go, stating the obvious:



On 27 Sep 2001, D. J. Bernstein wrote:

> Gustafsson's document says that caches ``SHOULD also decompress RRs of
> type RP, AFSDB, RT, SIG, PX, NXT, NAPTR, and SRV.''
>
> I objected to that text in December 2000, and I continue to object to
> it, for several reasons:
>
>    * RFC 1035 clearly prohibits compression except in owner names, NS
>      data, CNAME data, PTR data, MX data, and SOA data.

It does not. It only states that domain names in NS, SOA, CNAME, and PTR
rdata section may be compressed. Why ? Because the RDATA sections' format
is known. It does not prohibit RP AFSDB RT SIG PX NXT NAPTR and SRV rdata
sections' domain name compression. Why not ? Those did not exist at the
time.

>      Compression in
>      RP, AFSDB, etc. is inexcusable. There are no legitimate situations
>      where decompression of RP, AFSDB, etc. will accomplish anything.

RP's and PX's rdata contains two domain names. AFSDB, RT, SIG, NXT, NAPTR
and SRV's rdata contains a domain name. All can be compressed. There is no
documentation that prohibts it. Very legitimate. Very very legitimate
indeed. So, if its compressed, decompress it.

>    * My cache, which is deployed at tens of thousands of sites, handles
>      exactly the forms of compression allowed by RFC 1035.

I consider Claims with no substantial backing of real-world data are false
claims IMHO.

>      It does _not_ decompress RP, AFSDB, etc. It works just fine.

Hmm, how can that be. Luck I guess. Why luck ? Probably most pointers
point to the question sections QNAME. therefore in most cases that works.

>    * Gustafsson's text violates RFC 2119, section 6. It is not necessary
>      for interoperability. It is an abuse of the standards process.

It is necessary, therefor you are wrong.

> completely ignoring the substance of my objections.

For the obvious reasons.

> I also object to this document's failure to point out that transparent
> handling of new types is already required by the existing DNS standards.
> See http://cr.yp.to/djbdns/newtypes.html.

>From the NOTE WELL:

   Statements made outside of an IETF meeting, mailing list or other
   function, that are clearly not intended to be input to an IETF
   activity, group or function, are not subject to these provisions.

Regards,

Roy Arends
Nominum,
The BIND Company



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 11:16:13 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA03448
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 11:16:13 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mcbL-000AKy-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 07:57:23 -0700
Received: from [147.28.4.2] (helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mcbK-000AKs-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 07:57:23 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mcbI-0009Cv-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 07:57:20 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109271451.f8REpbr95227@drugs.dv.isc.org>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
From: Mark.Andrews@isc.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt 
In-reply-to: Your message of "27 Sep 2001 05:23:08 GMT."
             <20010927052308.691.qmail@cr.yp.to> 
Date: Fri, 28 Sep 2001 00:51:36 +1000
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

>    * BIND's support for bogus compressed records was exactly what led to
>      the http://www.cert.org/advisories/CA-2000-20.html disaster. We do
>      _not_ want more of these disasters.

	Dan,
	      I presume you are meaning the SRV bug.  If you understood
	the causes the bug you would know that it had nothing to do with
	whether the rdata was compressed or not other than it allowed a
	otherwise hidden bug to become visible.

	For the record the rdata of the SRV record was not compressed.

	Mark
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:25:25 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15254
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:25:24 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mk5f-000ITV-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 15:57:11 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mk5f-000ITP-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:57:11 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mk5e-0009OU-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:57:10 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 18:57:04 +0200
From: bert hubert <ahu@ds9a.nl>
To: Andreas Gustafsson <gson@isc.org>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: RCODE implementation done :-)
Message-ID: <20010927185704.A28699@outpost.ds9a.nl>
References: <20010927173604.A28105@outpost.ds9a.nl>
In-Reply-To: <20010927173604.A28105@outpost.ds9a.nl>; from ahu@ds9a.nl on Thu, Sep 27, 2001 at 05:36:04PM +0200
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Thu, Sep 27, 2001 at 05:36:04PM +0200, bert hubert wrote:

> We're now implementing RCODE, and one thing that comes to mind is what
> should 'win' in a zone:
> 
> tst	IN	RCODE1  4 02030405
> tst	IN	A	1.2.3.4

We're now 'done' implementing RCODE. The main thing we learned was that
'RCODE140' may be a tad long for many implementations, we had to stretch our
SQL tables to accomodate. You may want to use an enum construction
internally.

The entire diff is under 150 lines, so other implementors should not have a
lot of problems.

> The easiest way out is to specify that RCODE matches are only tried for
> records which are really unknown to the nameserver. However, this might lead
> to uncertainty. Another way out is to suggest that a nameserver should
> reject a zone with RCODE records for rrtypes it understands natively.

This is what we do now.

Regards,

bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
Netherlabs BV / Rent-a-Nerd.nl           - Nerd Available -
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:25:40 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15266
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:25:38 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mk1r-000IPf-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 15:53:15 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mk1r-000IPZ-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:53:15 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mk1q-0009O4-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:53:14 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109271608.f8RG8vr96452@drugs.dv.isc.org>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
From: Mark.Andrews@isc.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt 
In-reply-to: Your message of "27 Sep 2001 05:23:08 GMT."
             <20010927052308.691.qmail@cr.yp.to> 
Date: Fri, 28 Sep 2001 02:08:56 +1000
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Gustafsson's document says that caches ``SHOULD also decompress RRs of
> type RP, AFSDB, RT, SIG, PX, NXT, NAPTR, and SRV.''
> 
> I objected to that text in December 2000, and I continue to object to
> it, for several reasons:
> 
>    * RFC 1035 clearly prohibits compression except in owner names, NS
>      data, CNAME data, PTR data, MX data, and SOA data. Compression in
>      RP, AFSDB, etc. is inexcusable. There are no legitimate situations
>      where decompression of RP, AFSDB, etc. will accomplish anything.

	It will allow you to interoperate with implementations that
	followed the bad recommendations to compress the domainnames
	in some of the RFCs that describe this records.

>    * My cache, which is deployed at tens of thousands of sites, handles
>      exactly the forms of compression allowed by RFC 1035. It does _not_
>      decompress RP, AFSDB, etc. It works just fine.

	Provided you don't get answers from any of the above
	implementations.

> 
>    * Gustafsson's text violates RFC 2119, section 6. It is not necessary
>      for interoperability. It is an abuse of the standards process.

	Actually it is as we (the working group) stuffed up and put out
	RFCs that said you could compress what we shouldn't have.

>
> The text should be replaced by the opposite recommendation. Implementors
> should be encouraged to make their decompression code as limited as
> possible, handling RFC 1035 compression and nothing else.
> 
> Gustafsson's only response to these points has been ``it's a SHOULD, not
> a MUST,'' completely ignoring the substance of my objections.

	Well I'd make it MUST decompress or MUST detect compression
	and return SERVFAIL if found.  That we those that just want
	to make this work can be happy and those that feel they
	need to force people to upgrade can be happy.  In no case
	should they be sent on in compressed form.

	Mark
> 
> I also object to this document's failure to point out that transparent
> handling of new types is already required by the existing DNS standards.
> See http://cr.yp.to/djbdns/newtypes.html.
> 
> ---Dan
> 
> 
> to unsubscribe send a message to namedroppers-request@ops.ietf.org with
> the word 'unsubscribe' in a single line as the message text body.
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:27:01 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15297
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:27:00 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mk2i-000IQZ-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 15:54:08 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mk2i-000IQT-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:54:08 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mk2h-0009OD-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:54:07 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 09:21:32 -0700
From: Mark Gritter <mgritter@CS.Stanford.EDU>
To: Andreas Gustafsson <gson@isc.org>
Cc: Mark Gritter <mgritter@CS.Stanford.EDU>, namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
Message-ID: <20010927092131.A1348@Xenon.Stanford.EDU>
References: <E15m6zI-000NCA-00@rip.psg.com> <20010926144239.A22605@Xenon.Stanford.EDU> <200109262226.PAA03883@gulag.araneus.fi>
In-Reply-To: <200109262226.PAA03883@gulag.araneus.fi>; from gson@isc.org on Wed, Sep 26, 2001 at 03:26:13PM -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Wed, Sep 26, 2001 at 03:26:13PM -0700, Andreas Gustafsson wrote:
> This is just not how caching servers work.  Caching servers don't
> cache entire responses, they cache individual resource record sets.
> What you are proposing is not a change to the unknown-rrs draft but
> a fundamental change to how the DNS works.

I realize that this is the case for caching servers, and I do not wish
to force such a change.  As I pointed out a little later, the best thing to 
do with messages such as the ones I envision might be to not cache the
RRs at all (which can be done by setting TTL=0 and doesn't require 
software changes.)

My concern is primarily with forwarding behavior.

> 
> There is no such thing as an end-to-end protocol extension in the DNS.
> All existing extension that rely on special record types (EDNS0, TSIG,
> SIG0) are hop-by-hop, i.e., they don't traverse caching servers.

That does not imply that there cannot or should not be such extensions.

The change I am asking for is a requirement that unknown records 
not be dropped from either requests or responses.  This need not 
apply to answers generated from a cache.

This does require that caching DNS servers build their (non-cached) responses 
directly from the answers they receive rather than putting the RRs into a 
cache and then synthesizing the result.  It is not hard, in my experience, 
to do so, since the necessary mapping between pending responses and client 
requests already exists.

I admit that other decisions might be the technically correct thing to do.  

However, it will be impossible to deploy any future extensions without
defining explicitly what a recursive name server is supposed to do with
unknown RRs, and "transparently" doesn't cut it.  So, stepping back
for a moment, I think there _is_ behavior that needs to be specified.

If a query contains unknown RRs, is the correct behavior
  1. Ignore them
  2. Send back SERVFAIL (seen in the wild) or FORMERR
  3. Pass them along if no answer found in the cache
  4. Always forward the query rather than looking in the cache

(Some existing DNS servers return SERVFAIL if the query _type_ is unknown...)

Similarly, if a reponse contains unknown RRs, is the correct behavior
  1. Pass just the answers back to the original requester
  2. Pass both answers and additional records back to the original requester
  3. Pass back only CNAME records and those matching QTYPE, doing own
     additional section processing?

If some of the response types do not match QTYPE, what is the proper
"transparent" behavior?

Even if we don't agree on what transparent _should_ mean, I hope that
we can come up with some language that at least allows designers to
know more specifically what to expect.

-- 
Mark Gritter <mgritter@cs.stanford.edu>
http://www-cs-students.stanford.edu/~mgritter/


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:27:11 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15313
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:27:10 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mk27-000IPo-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 15:53:31 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mk26-000IPi-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:53:30 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mk25-0009OA-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:53:29 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109271613.f8RGDar96471@drugs.dv.isc.org>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
From: Mark.Andrews@isc.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt 
In-reply-to: Your message of "27 Sep 2001 05:23:08 GMT."
             <20010927052308.691.qmail@cr.yp.to> 
Date: Fri, 28 Sep 2001 02:13:35 +1000
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> I also object to this document's failure to point out that transparent
> handling of new types is already required by the existing DNS standards.

	The existing RFC's failed to cover master file formats for
	unknown types.  As such they were incomplete.

>From RFC 1034.

The standard format of master files allows them to be exchanged between
hosts (via FTP, mail, or some other mechanism); this facility is useful
when an organization wants a domain, but doesn't want to support a name
server.  The organization can maintain the master files locally using a 
text editor, transfer them to a foreign host which runs a name server,
and then arrange with the system administrator of the name server to get
the files loaded.

	Mark


--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:27:38 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15324
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:27:37 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mjzH-000IM5-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 15:50:35 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mjzG-000ILy-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:50:34 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mjzA-0009NX-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 15:50:28 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 17:36:04 +0200
From: bert hubert <ahu@ds9a.nl>
To: Andreas Gustafsson <gson@isc.org>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: should RCODE override a regular record in the zone?
Message-ID: <20010927173604.A28105@outpost.ds9a.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

We're now implementing RCODE, and one thing that comes to mind is what
should 'win' in a zone:

tst	IN	RCODE1  4 02030405
tst	IN	A	1.2.3.4

Now a request comes along for the A record of tst, what should be in the
answer section? 

This is an admittedly trivial example, but some guidence would be nice. This
involves modifying the RFC1034 algorithm.

The easiest way out is to specify that RCODE matches are only tried for
records which are really unknown to the nameserver. However, this might lead
to uncertainty. Another way out is to suggest that a nameserver should
reject a zone with RCODE records for rrtypes it understands natively.

The clearest solution, which does require real work on the part of
nameserver implementors is to treat RCODE and 'native' equivalently, which
would in this case lead to an answer of '1.2.3.4' AND '2.3.4.5'.

Sorry for butting in this late, but non-specification of this behaviour will
lead to serious incompatibilities, which was just what we tried to prevent.

Regards,

bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
Netherlabs BV / Rent-a-Nerd.nl           - Nerd Available -
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:27:41 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15335
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:27:39 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mk8e-000IW5-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:00:16 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mk8e-000IVz-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:00:16 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mk8d-0009Oq-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:00:15 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: 27 Sep 2001 17:19:24 -0000
Message-ID: <20010927171924.12654.qmail@cr.yp.to>
From: "D. J. Bernstein" <djb@cr.yp.to>
To: namedroppers@ops.ietf.org
Subject: Bogus compression is already prohibited
References: <20010927052308.691.qmail@cr.yp.to> <Pine.BSF.4.33.0109271507140.12760-100000@node10c4d.a2000.nl>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

The existing DNS standards already state clearly that caches should save
RRs of unknown types. This is, as an engineering matter, obviously
incompatible with servers that compress new types.

Inescapable conclusion: Servers must not compress new types.

Roy Arends writes:
> D. J. Bernstein wrote:
> > * RFC 1035 clearly prohibits compression except in owner names, NS
> >   data, CNAME data, PTR data, MX data, and SOA data.
> It does not.

RFC 1035 says

   Since the domain system is intentionally extensible, new data types
   ... should always be expected. ...

   Pointers can only be used for occurances [sic] of a domain name where
   the format is not class specific. If this were not the case, a name
   server or resolver would be required to know the format of all RRs it
   handled.

In case that isn't clear enough: RFC 1123 says

   DNS software ... SHOULD be written to minimize the trauma associated
   with the introduction of new well-known types and local
   experimentation with non-standard types. ...

   Compression relies on knowledge of the format of data inside a
   particular RR. Hence compression must only be used for the contents
   of well-known, class-independent RRs, and must never be used for
   class-specific RRs or RR types that are not well-known. ...

   A resolver can receive [unrecognized data types] as the result of
   queries. For proper operation, this data must be preserved.

Pretty damn clear.

> RP's and PX's rdata contains two domain names. AFSDB, RT, SIG, NXT, NAPTR
> and SRV's rdata contains a domain name. All can be compressed. There is no
> documentation that prohibts it. Very legitimate. Very very legitimate
> indeed. So, if its compressed, decompress it.

Wrong. As I said, compression is RP, AFSDB, etc. is inexcusable. BIND
hasn't screwed this up since version 8.1.2, right?

---Dan


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:30:19 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15380
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:30:16 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mkGi-000Igu-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:08:36 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mkGh-000Ign-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:08:36 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mkGh-0009PH-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:08:35 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: 27 Sep 2001 17:45:57 -0000
Message-ID: <20010927174557.30969.qmail@cr.yp.to>
From: "D. J. Bernstein" <djb@cr.yp.to>
To: namedroppers@ops.ietf.org
Subject: Minimize security-critical code
References: <20010927052308.691.qmail@cr.yp.to> <200109271451.f8REpbr95227@drugs.dv.isc.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Mark.Andrews@isc.org writes:
> If you understood the causes the bug you would know that it had
> nothing to do with whether the rdata was compressed or not

According to a bugtraq message, the SRV bug was ``a bug in the handling
of the compression pointer tables'' when a BIND 8.2.2-P6 cache received
certain types of SRV records.

This disaster would never have happened if BIND had simply treated SRV
as opaque data.

---Dan


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:32:46 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15411
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:32:44 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mkF0-000IeR-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:06:50 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mkEz-000IeD-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:06:49 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mkEy-0009P7-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:06:48 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 10:42:45 -0700 (PDT)
Message-Id: <200109271742.KAA04466@gulag.araneus.fi>
To: Kevin Darcy <kcd@daimlerchrysler.com>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <3BB251E7.E77BE8AB@daimlerchrysler.com>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926111745.A10616@outpost.ds9a.nl>
	<200109261904.MAA03755@gulag.araneus.fi>
	<3BB23A32.4FB62E0D@daimlerchrysler.com>
	<200109262147.OAA03864@gulag.araneus.fi>
	<3BB251E7.E77BE8AB@daimlerchrysler.com>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Kevin Darcy writes:
> > I doubt the benefits of additional compression will outweigh the cost
> > of such a mechanism.  But if the working group does decide to
> > standardize a mechanism for communicating extended compression support
> > from client to server, its specification can serve as an update to the
> > unknown-rrs specification, modifying the rules for where compression
> > is allowed.  There is no need to change the text of the unknown-rrs
> > draft itself in anticipation of this.
> 
> Sorry, I didn't make myself very clear. I wasn't referring to a generic "extended
> compression support" negotiation feature, I simply meant that if a new record type
> is only going to be used between clients and servers who are already known to
> understand the type, e.g. the server only sends records of that type if the client
> sent an EDNS0 OPTION requesting their inclusion, which the server understands and
> with which it complies, then I see no reason why they can't use name compression in
> their RDATA. Such "negotiated" compression is forbidden under the current draft
> wording, and I find that too rigid.

Even then, I would prefer that exceptions to the "no compression" rule
be stated in the proposals defining the RR types and/or EDNS0 options
they apply to, to ensure that their consequences get carefully weighed
on a case-by-case basis.  Including a blanket endorsement of
negotiated compression schemes in the unknown-rrs draft itself would
only encourage their indiscriminate use.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:40:46 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15469
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:40:45 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mkLK-000Inl-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:13:22 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mkLJ-000Inf-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:13:21 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mkLI-0009Pa-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:13:20 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 11:29:14 -0700 (PDT)
Message-Id: <200109271829.LAA04490@gulag.araneus.fi>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010927052308.691.qmail@cr.yp.to>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010927052308.691.qmail@cr.yp.to>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

D. J. Bernstein writes:
> Gustafsson's document says that caches ``SHOULD also decompress RRs of
> type RP, AFSDB, RT, SIG, PX, NXT, NAPTR, and SRV.''
> 
> I objected to that text in December 2000, and I continue to object to
> it, for several reasons:
> 
>    * RFC 1035 clearly prohibits compression except in owner names, NS
>      data, CNAME data, PTR data, MX data, and SOA data.

RFC1035 neither clearly allows nor clearly prohibits compression in
future types.  It only touches on the subject in the following
paragraph:

  Pointers can only be used for occurances of a domain name where the
  format is not class specific.  If this were not the case, a name server
  or resolver would be required to know the format of all RRs it handled.
  As yet, there are no such cases, but they may occur in future RDATA
  formats.

This is confusing and self-contradictory.  It implies that future
non-class-specific RRs can be compressed without servers or resolvers
knowing the format of all RRs, but as we all know, that is not
actually the case.

>      Compression in RP, AFSDB, etc. is inexcusable.

I fully agree.  That is why the draft says it MUST NOT be done.

>      There are no legitimate situations
>      where decompression of RP, AFSDB, etc. will accomplish anything.

It will accomplish interoperability with existing broken implementations
that are in widespread use.

>    * Gustafsson's text violates RFC 2119, section 6. It is not necessary
>      for interoperability. It is an abuse of the standards process.

Even interoperating with common broken implementations counts as
interoperability.

> I also object to this document's failure to point out that transparent
> handling of new types is already required by the existing DNS standards.

I don't have a problem with adding a reference to RFC1123.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:42:02 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15493
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:42:00 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mkRW-000Ixg-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:19:46 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mkRW-000IxY-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:19:46 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mkRV-0009QJ-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:19:45 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 12:29:12 -0700 (PDT)
Message-Id: <200109271929.MAA04519@gulag.araneus.fi>
To: Mark Gritter <mgritter@CS.Stanford.EDU>
Cc: namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <20010927092131.A1348@Xenon.Stanford.EDU>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926144239.A22605@Xenon.Stanford.EDU>
	<200109262226.PAA03883@gulag.araneus.fi>
	<20010927092131.A1348@Xenon.Stanford.EDU>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Mark Gritter writes:
> On Wed, Sep 26, 2001 at 03:26:13PM -0700, Andreas Gustafsson wrote:
> > This is just not how caching servers work.  Caching servers don't
> > cache entire responses, they cache individual resource record sets.
> > What you are proposing is not a change to the unknown-rrs draft but
> > a fundamental change to how the DNS works.
> 
> I realize that this is the case for caching servers, and I do not wish
> to force such a change.  As I pointed out a little later, the best thing to 
> do with messages such as the ones I envision might be to not cache the
> RRs at all (which can be done by setting TTL=0 and doesn't require 
> software changes.)
> 
> My concern is primarily with forwarding behavior.

You seem to be making a distinction between caching and forwarding
servers that doesn't really exist.  In existing caching servers,
forwarding is just a way to override the mechanism for determining
which servers to ask.  The response to a forwarded query is processed
and cached just like the response to any other query.

> > There is no such thing as an end-to-end protocol extension in the DNS.
> > All existing extension that rely on special record types (EDNS0, TSIG,
> > SIG0) are hop-by-hop, i.e., they don't traverse caching servers.
> 
> That does not imply that there cannot or should not be such extensions.
>
> The change I am asking for is a requirement that unknown records 
> not be dropped from either requests or responses.

The intent of the draft is that servers should treat records of
unknown type the same way it treats records of known type.  Existing
servers will drop records of known type that are irrelevant to the
query (e.g., BIND 9 would always drop an unexpected PTR in the
additional section of a response to an A query, and BIND 8 would drop
it when answering from the cache), and I expect them to treat records
of unknown type the same way.  The treatment of unexpected records in
queries and responses (whether of known or unknown type) may be poorly
specified in the existing DNS standards, but that issue is orthogonal
to the issue the draft is trying to address.

Besides, your hypothetical end-to-end extension should allocate its
special RR type from the range reserved for meta-types in RFC2929,
which is specifically excluded from the definition of "unknown type"
in the unknown-rrs draft.  Therefore, the treatment of such records is
outside the scope of the draft.  If you wish to propose a standard way
of handling unknown meta-records, I suggest you do that as a separate
draft.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:44:33 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15542
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:44:32 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mkVR-000J5K-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:23:49 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mkVR-000J5E-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:23:49 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mkVQ-0009QR-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:23:48 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 12:37:06 -0700 (PDT)
Message-Id: <200109271937.MAA04525@gulag.araneus.fi>
To: bert hubert <ahu@ds9a.nl>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: should RCODE override a regular record in the zone?
In-Reply-To: <20010927173604.A28105@outpost.ds9a.nl>
References: <20010927173604.A28105@outpost.ds9a.nl>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

bert hubert writes:
> We're now implementing RCODE, and one thing that comes to mind is what
> should 'win' in a zone:
> 
> tst	IN	RCODE1  4 02030405
> tst	IN	A	1.2.3.4

RCODE1?  Do you mean TYPE1?  Also, you are missing the \# marker.

> Now a request comes along for the A record of tst, what should be in the
> answer section?

A resource record set consisting of two A records.  The unknown-rrs
draft says:

   Even though an RR of known type represented in the \# format is
   effectively treated as an unknown type for the purpose of parsing the
   RDATA text representation, all further processing by the server MUST
   treat it as a known type [...]

In other words, your example is treated exactly the same way as if
the zone had contained two ordinary A records.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 19:50:14 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA15643
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 19:50:13 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mkbD-000JEc-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:29:47 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mkbC-000JEW-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:29:47 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mkbC-0009Qf-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:29:46 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 22:35:32 +0200
From: bert hubert <ahu@ds9a.nl>
To: Andreas Gustafsson <gson@isc.org>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: should RCODE override a regular record in the zone?
Message-ID: <20010927223532.A29890@outpost.ds9a.nl>
References: <20010927173604.A28105@outpost.ds9a.nl> <200109271937.MAA04525@gulag.araneus.fi>
In-Reply-To: <200109271937.MAA04525@gulag.araneus.fi>; from gson@isc.org on Thu, Sep 27, 2001 at 12:37:06PM -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Thu, Sep 27, 2001 at 12:37:06PM -0700, Andreas Gustafsson wrote:

> A resource record set consisting of two A records.  The unknown-rrs
> draft says:
> 
>    Even though an RR of known type represented in the \# format is
>    effectively treated as an unknown type for the purpose of parsing the
>    RDATA text representation, all further processing by the server MUST
>    treat it as a known type [...]
> 
> In other words, your example is treated exactly the same way as if
> the zone had contained two ordinary A records.

Ok, missed that. For a nameserver that parses its zones on startup, this
means some extra work once (it needs to flag known-unknown rrs). 

For nameservers that work from a database, this means an additional lookup
for *each* query to figure out if there are 'known-unknown rrs' around.

In fact, it might be better to make each query an ANY query internally
then.

Regards,


bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
Netherlabs BV / Rent-a-Nerd.nl           - Nerd Available -
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Thu Sep 27 20:09:55 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA16002
	for <dnsext-archive@lists.ietf.org>; Thu, 27 Sep 2001 20:09:52 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15mknT-000Ji0-00
	for namedroppers-data@psg.com; Thu, 27 Sep 2001 16:42:27 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15mknS-000Jhu-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:42:26 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15mknR-0009RA-00
	for namedroppers@ops.ietf.org; Thu, 27 Sep 2001 16:42:25 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BB3A0A7.3B63FB1B@daimlerchrysler.com>
Date: Thu, 27 Sep 2001 17:56:55 -0400
From: Kevin Darcy <kcd@daimlerchrysler.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <E15m6zI-000NCA-00@rip.psg.com>
		<20010926111745.A10616@outpost.ds9a.nl>
		<200109261904.MAA03755@gulag.araneus.fi>
		<3BB23A32.4FB62E0D@daimlerchrysler.com>
		<200109262147.OAA03864@gulag.araneus.fi>
		<3BB251E7.E77BE8AB@daimlerchrysler.com> <200109271742.KAA04466@gulag.araneus.fi>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

gson@isc.org wrote:

> Kevin Darcy writes:
> > > I doubt the benefits of additional compression will outweigh the cost
> > > of such a mechanism.  But if the working group does decide to
> > > standardize a mechanism for communicating extended compression support
> > > from client to server, its specification can serve as an update to the
> > > unknown-rrs specification, modifying the rules for where compression
> > > is allowed.  There is no need to change the text of the unknown-rrs
> > > draft itself in anticipation of this.
> >
> > Sorry, I didn't make myself very clear. I wasn't referring to a generic "extended
> > compression support" negotiation feature, I simply meant that if a new record type
> > is only going to be used between clients and servers who are already known to
> > understand the type, e.g. the server only sends records of that type if the client
> > sent an EDNS0 OPTION requesting their inclusion, which the server understands and
> > with which it complies, then I see no reason why they can't use name compression in
> > their RDATA. Such "negotiated" compression is forbidden under the current draft
> > wording, and I find that too rigid.
>
> Even then, I would prefer that exceptions to the "no compression" rule
> be stated in the proposals defining the RR types and/or EDNS0 options
> they apply to, to ensure that their consequences get carefully weighed
> on a case-by-case basis.  Including a blanket endorsement of
> negotiated compression schemes in the unknown-rrs draft itself would
> only encourage their indiscriminate use.

I think that, without some sort of broad exception like that, the compression
prohibition oversteps the self-announced charter of the draft, inasmuch as it prohibits
compression even in "known" RRs ("known" to all relevant parties, that is, i.e. the
client and server), and basically declares all new record types guilty until proven
innocent in this regard. If you want to sweep that broadly, you should rename the draft
to "Rules for Addition of New Record Types" or something similar.


- Kevin





to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 13:41:47 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA20681
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:41:44 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1CU-000BSl-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:13:22 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1CT-000BSe-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:13:21 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1CT-0009XU-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:13:21 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Thu, 27 Sep 2001 19:57:26 -0700 (PDT)
From: Brian Wellington <Brian.Wellington@nominum.com>
To: bert hubert <ahu@ds9a.nl>
cc: Andreas Gustafsson <gson@isc.org>,
        namedroppers <namedroppers@ops.ietf.org>
Subject: Re: should RCODE override a regular record in the zone?
In-Reply-To: <20010927223532.A29890@outpost.ds9a.nl>
Message-ID: <Pine.LNX.4.33.0109271950160.17618-100000@spratly.nominum.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Thu, 27 Sep 2001, bert hubert wrote:

> On Thu, Sep 27, 2001 at 12:37:06PM -0700, Andreas Gustafsson wrote:
>
> > A resource record set consisting of two A records.  The unknown-rrs
> > draft says:
> >
> >    Even though an RR of known type represented in the \# format is
> >    effectively treated as an unknown type for the purpose of parsing the
> >    RDATA text representation, all further processing by the server MUST
> >    treat it as a known type [...]
> >
> > In other words, your example is treated exactly the same way as if
> > the zone had contained two ordinary A records.
>
> Ok, missed that. For a nameserver that parses its zones on startup, this
> means some extra work once (it needs to flag known-unknown rrs).

Why?  When parsing the zone, it parses each record regardless of whether
it is known or unknown.

> For nameservers that work from a database, this means an additional lookup
> for *each* query to figure out if there are 'known-unknown rrs' around.

Again, why?  The database should have a simple type number -> rrset
mapping.  There's no fundamental difference whether the data was loaded
from the canonical format or the unknown rr format.  Database lookups have
nothing to do with the textual representation of the type code.

If, however, you are storing data in a database that stores the textual
representation of types, you should convert unknown RR type formatted
records into their canonical form before storing them, when the type is
known.  For example, if you see a TYPE1 record, store it as an A record.

> In fact, it might be better to make each query an ANY query internally
> then.

No, this would be an exceedingly bad idea, and is completely unnecessary.

Brian



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 13:41:47 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA20684
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:41:44 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n17S-000BKT-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:08:10 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n17S-000BKN-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:08:10 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n17R-0009XL-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:08:09 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109280159.f8S1xtr99782@drugs.dv.isc.org>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
From: Mark.Andrews@isc.org
Subject: Re: Minimize security-critical code 
In-reply-to: Your message of "27 Sep 2001 17:45:57 GMT."
             <20010927174557.30969.qmail@cr.yp.to> 
Date: Fri, 28 Sep 2001 11:59:54 +1000
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Mark.Andrews@isc.org writes:
> > If you understood the causes the bug you would know that it had
> > nothing to do with whether the rdata was compressed or not
> 
> According to a bugtraq message, the SRV bug was ``a bug in the handling
> of the compression pointer tables'' when a BIND 8.2.2-P6 cache received
> certain types of SRV records.
> 
> This disaster would never have happened if BIND had simply treated SRV
> as opaque data.
> 
> ---Dan

	Dan,
	      I'd describe how to trigger the bug in detail except that
	then people would be deliberating DoS the existing servers that
	havn't upgraded yet.

	I fixed the bug and had to workout which version of BIND were
	affected.  The bug exists in more versions that those listed
	on http://www.isc.org/products/BIND/bind-security.html but
	it cannot be triggered in the earlier versions.  Only the
	version listed can be DoS'd.  The bug predates SRV by many
	years.

	Mark
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 13:41:49 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA20682
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:41:44 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1PU-000BlL-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:26:48 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1PT-000BlE-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:26:47 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1PO-0009YI-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:26:42 -0700
Date: Fri, 28 Sep 2001 10:44:22 +0200
From: bert hubert <ahu@ds9a.nl>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: should RCODE override a regular record in the zone?
Message-ID: <20010928104422.A3283@outpost.ds9a.nl>
References: <20010927223532.A29890@outpost.ds9a.nl> <Pine.LNX.4.33.0109271950160.17618-100000@spratly.nominum.com>
Mime-Version: 1.0
In-Reply-To: <Pine.LNX.4.33.0109271950160.17618-100000@spratly.nominum.com>; from Brian.Wellington@nominum.com on Thu, Sep 27, 2001 at 07:57:26PM -0700
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

On Thu, Sep 27, 2001 at 07:57:26PM -0700, Brian Wellington wrote:

> If, however, you are storing data in a database that stores the textual
> representation of types, you should convert unknown RR type formatted
> records into their canonical form before storing them, when the type is
> known.  For example, if you see a TYPE1 record, store it as an A record.

The main reason for being smart about 'known unknown types' is upgrades.
Only then will you encounter a TYPE# which you now understand natively. 

> > In fact, it might be better to make each query an ANY query internally
> > then.
> 
> No, this would be an exceedingly bad idea, and is completely unnecessary.

I suggest you write a database driven nameserver before making statements
about them.

Anyway, we've implemented it because it is useful for us - it provides a lot
of flexibility because of the ability to serve unknown types. I'm not sure
if we'll follow the 'known unknown types are equivalent with natice types'
rule.

Regards,

bert

-- 
http://www.PowerDNS.com          Versatile DNS Software & Services
Trilab                                 The Technology People
Netherlabs BV / Rent-a-Nerd.nl           - Nerd Available -
'SYN! .. SYN|ACK! .. ACK!' - the mating call of the internet


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 13:43:21 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA20764
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:43:16 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1Rv-000Bpt-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:29:19 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1Ru-000Bpm-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:29:19 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1Rt-0009Ya-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:29:17 -0700
Message-Id: <200109281109.HAA09046@ietf.org>
Mime-Version: 1.0
Content-Type: Multipart/Mixed; Boundary="NextPart"
To: IETF-Announce: ;
Cc: namedroppers@ops.ietf.org
From: Internet-Drafts@ietf.org
Reply-to: Internet-Drafts@ietf.org
Subject: I-D ACTION:draft-ietf-dnsext-ipv6-addresses-00.txt
Date: Fri, 28 Sep 2001 07:09:25 -0400
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the DNS Extensions Working Group of the IETF.

	Title		: Representing IPv6 addresses in DNS
	Author(s)	: R. Bush, A. Durand, B. Fink, O. Gudmundsson, T. Hain
	Filename	: draft-ietf-dnsext-ipv6-addresses-00.txt
	Pages		: 6
	Date		: 27-Sep-01
	
This document clarifies and updates the standards status of RFCs that
define direct and reverse map of IPv6 addresses in DNS. This document
moves the A6 and Bit label specifications to experimental status.

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-dnsext-ipv6-addresses-00.txt

To remove yourself from the IETF Announcement list, send a message to 
ietf-announce-request with the word unsubscribe in the body of the message.

Internet-Drafts are also available by anonymous FTP. Login with the username
"anonymous" and a password of your e-mail address. After logging in,
type "cd internet-drafts" and then
	"get draft-ietf-dnsext-ipv6-addresses-00.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt


Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-dnsext-ipv6-addresses-00.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.
		
		
Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body;
	access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID:	<20010927141530.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-dnsext-ipv6-addresses-00.txt

--OtherAccess
Content-Type: Message/External-body;
	name="draft-ietf-dnsext-ipv6-addresses-00.txt";
	site="ftp.ietf.org";
	access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID:	<20010927141530.I-D@ietf.org>

--OtherAccess--

--NextPart--




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 13:51:54 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA21054
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:51:53 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1Vq-000Bvb-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:33:22 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1Vp-000BvU-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:33:22 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1Vg-0009Yv-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:33:12 -0700
Date: Fri, 28 Sep 2001 16:44:52 +0300 (EEST)
From: Pekka Savola <pekkas@netcore.fi>
To: <namedroppers@ops.ietf.org>
Subject: Re: I-D ACTION:draft-ietf-dnsext-ipv6-addresses-00.txt
In-Reply-To: <200109281109.HAA09046@ietf.org>
Message-ID: <Pine.LNX.4.33.0109281643370.9190-100000@netcore.fi>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk

On Fri, 28 Sep 2001 Internet-Drafts@ietf.org wrote:
> 	Title		: Representing IPv6 addresses in DNS
> 	Author(s)	: R. Bush, A. Durand, B. Fink, O. Gudmundsson, T. Hain
> 	Filename	: draft-ietf-dnsext-ipv6-addresses-00.txt
> 	Pages		: 6
> 	Date		: 27-Sep-01

Just a note; I support this but in all honesty, perhaps s/consensus/rough
consensus/ would be closer to the community opinion. :-)

-- 
Pekka Savola                 "Tell me of difficulties surmounted,
Netcore Oy                   not those you stumble over and fall"
Systems. Networks. Security.  -- Robert Jordan: A Crown of Swords



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 13:58:19 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA21229
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:58:16 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1bv-000C73-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:39:39 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1bu-000C6w-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:39:39 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1VZ-0009Yt-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:33:05 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Fri, 28 Sep 2001 08:42:06 -0500
From: Matt Crawford <crawdad@fnal.gov>
Subject: Re: Bogus compression is already prohibited
In-reply-to: "27 Sep 2001 17:19:24 -0000." <20010927171924.12654.qmail@cr.yp.to>
To: namedroppers@ops.ietf.org
Message-id: <200109281342.f8SDg6m22765@gungnir.fnal.gov>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> RFC 1035 says
>    Pointers can only be used for occurances [sic] of a domain name where
>    the format is not class specific. 

And this (class-independence) was clearly the wrong condition.

> In case that isn't clear enough: RFC 1123 says
> [...]

Elsewhere RFC 1123 says
         At every layer of the protocols, there is a general rule whose
         application can lead to enormous benefits in robustness and
         interoperability:

                "Be liberal in what you accept, and
                 conservative in what you send"

         Software should be written to deal with every conceivable
         error, no matter how unlikely; sooner or later a packet will
         come in with that particular combination of errors and
         attributes, and unless the software is prepared, chaos can
         ensue.  In general, it is best to assume that the network is
         filled with malevolent entities that will send in packets
         designed to have the worst possible effect.  This assumption
         will lead to suitable protective design, although the most
         serious problems in the Internet have been caused by
         unenvisaged mechanisms triggered by low-probability events;
         mere human malice would never have taken so devious a course!

         Adaptability to change must be designed into all levels of
         Internet host software.  As a simple example, consider a
         protocol specification that contains an enumeration of values
         for a particular header field -- e.g., a type field, a port
         number, or an error code; this enumeration must be assumed to
         be incomplete.  Thus, if a protocol specification defines four
         possible error codes, the software must not break when a fifth
         code shows up.  An undefined code might be logged (see below),
         but it must not cause a failure.

As you say, "Pretty damn clear".


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 14:11:03 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA21473
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 14:11:01 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1kD-000CM7-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:48:13 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1kC-000CM1-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:48:12 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1kC-0009aF-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:48:12 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Fri, 28 Sep 2001 10:44:22 -0700 (PDT)
Message-Id: <200109281744.KAA05226@gulag.araneus.fi>
To: Kevin Darcy <kcd@daimlerchrysler.com>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <3BB3A0A7.3B63FB1B@daimlerchrysler.com>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926111745.A10616@outpost.ds9a.nl>
	<200109261904.MAA03755@gulag.araneus.fi>
	<3BB23A32.4FB62E0D@daimlerchrysler.com>
	<200109262147.OAA03864@gulag.araneus.fi>
	<3BB251E7.E77BE8AB@daimlerchrysler.com>
	<200109271742.KAA04466@gulag.araneus.fi>
	<3BB3A0A7.3B63FB1B@daimlerchrysler.com>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> I think that, without some sort of broad exception like that, the
> compression prohibition oversteps the self-announced charter of the
> draft, inasmuch as it prohibits compression even in "known" RRs
> ("known" to all relevant parties, that is, i.e. the client and
> server), and basically declares all new record types guilty until
> proven innocent in this regard. If you want to sweep that broadly, you
> should rename the draft to "Rules for Addition of New Record Types" or
> something similar.

I disagree.  The prohibition on compression is not a new requirement.
The draft is only documenting an existing, though poorly documented,
restriction.

Even negotiated compression should be considered unsafe until proven
safe.  As a hypothetical example of what it might break, consider a
DNS translator unit built into a NAT or firewall modifying DNS packets
on the fly.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 14:29:40 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA20683
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 13:41:44 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n1Ke-000Bdd-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 10:21:48 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n1Kd-000BdW-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:21:47 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n1KX-0009Y0-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 10:21:41 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Robert Elz <kre@munnari.OZ.AU>
To: Kevin Darcy <kcd@daimlerchrysler.com>
cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt 
In-Reply-To: <3BB3A0A7.3B63FB1B@daimlerchrysler.com> 
References: <3BB3A0A7.3B63FB1B@daimlerchrysler.com>  <E15m6zI-000NCA-00@rip.psg.com> <20010926111745.A10616@outpost.ds9a.nl> <200109261904.MAA03755@gulag.araneus.fi> <3BB23A32.4FB62E0D@daimlerchrysler.com> <200109262147.OAA03864@gulag.araneus.fi> <3BB251E7.E77BE8AB@daimlerchrysler.com> <200109271742.KAA04466@gulag.araneus.fi> 
Date: Fri, 28 Sep 2001 13:22:23 +0700
Message-ID: <1376.1001658143@brandenburg.cs.mu.OZ.AU>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

    Date:        Thu, 27 Sep 2001 17:56:55 -0400
    From:        Kevin Darcy <kcd@daimlerchrysler.com>
    Message-ID:  <3BB3A0A7.3B63FB1B@daimlerchrysler.com>

  | inasmuch as it prohibits compression even in "known" RRs

Didn't we once have at least an agreement in principle that some later
EDNS version could degree that RR types <list of types> are "well known"
to implementations that understand that EDNS version, and hence compression
would be allowed in the RDATA for those RR types?

The EDNS doc would need to explicitly say which RRs it was including.

That seems to me to be the right way to gradually permit compression in
more RR types - EDNS negotiation is point to point, so it works in all
the right cases (which is also needed for packet length changes, etc).

Doing it this way (even using a flag bit to say "the compression part of
EDNSn is implemented" if that seems better than just "must be implemented")
provides the negotiation mechanism to allow sender & recipient (and
any intervening caches) all know what is going on, or avoid the issue.

What's more, as that would be a later RFC than this one, it can update this
one without caring what the current doc says must or must not be done.

This doesn't sound to me like an issue that we really need to be wasting
time on now.

kre



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 17:59:55 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA25744
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 17:59:53 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n5RI-000I1r-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 14:44:56 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n5RH-000I1l-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:44:56 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n5RB-0009fS-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:44:49 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: namedroppers@ops.ietf.org
Subject: re: I-D ACTION:draft-ietf-dnsext-ipv6-addresses-00.txt
From: Paul Vixie <vixie@vix.com>
Date: 28 Sep 2001 12:42:32 -0700
Message-ID: <g3elordrvb.fsf@as.vix.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Pekka Savola <pekkas@netcore.fi> writes:

> Just a note; I support this but in all honesty, perhaps s/consensus/rough
> consensus/ would be closer to the community opinion. :-)

indeed, that is true.  i am a strong dissenter.  A6 has to be mandatory if
IPv6 is going to solve more problems than it creates.

by the way, BIND 9.2 has AAAA synthesis, so recursive clients don't need to
follow A6 chains or even know that A6 exists.  so A6's experimentality might
end up being as irrelevant as, say, SRV's experimentality.  here's hoping.


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 17:59:57 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA25745
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 17:59:55 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n5ST-000I3Z-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 14:46:09 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n5SS-000I3T-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:46:08 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n5SM-0009fZ-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:46:02 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BB4E766.11981711@daimlerchrysler.com>
Date: Fri, 28 Sep 2001 17:11:02 -0400
From: Kevin Darcy <kcd@daimlerchrysler.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <E15m6zI-000NCA-00@rip.psg.com>
		<20010926111745.A10616@outpost.ds9a.nl>
		<200109261904.MAA03755@gulag.araneus.fi>
		<3BB23A32.4FB62E0D@daimlerchrysler.com>
		<200109262147.OAA03864@gulag.araneus.fi>
		<3BB251E7.E77BE8AB@daimlerchrysler.com>
		<200109271742.KAA04466@gulag.araneus.fi>
		<3BB3A0A7.3B63FB1B@daimlerchrysler.com> <200109281744.KAA05226@gulag.araneus.fi>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

gson@isc.org wrote:

> > I think that, without some sort of broad exception like that, the
> > compression prohibition oversteps the self-announced charter of the
> > draft, inasmuch as it prohibits compression even in "known" RRs
> > ("known" to all relevant parties, that is, i.e. the client and
> > server), and basically declares all new record types guilty until
> > proven innocent in this regard. If you want to sweep that broadly, you
> > should rename the draft to "Rules for Addition of New Record Types" or
> > something similar.
>
> I disagree.  The prohibition on compression is not a new requirement.
> The draft is only documenting an existing, though poorly documented,
> restriction.

If all you're doing is "documenting" requirements, shouldn't you at least
be noting the fact that the prohibition -- or at least the rationale for
the prohibition -- is partially obsoleted by EDNS0? If you don't note that,
then I'd say your documentation is incomplete.

> Even negotiated compression should be considered unsafe until proven
> safe.  As a hypothetical example of what it might break, consider a
> DNS translator unit built into a NAT or firewall modifying DNS packets
> on the fly.

Then turn the extension OFF if you're behind one of those beasts! That's
one of the main strengths of negotiated extensions -- they can always be
turned off if talking to/through something that doesn't understand them.


- Kevin





to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 18:01:06 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA25811
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 18:01:05 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n5Qw-000I1d-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 14:44:34 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n5Qv-000I1X-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:44:33 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n5Qp-0009fP-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:44:27 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Fri, 28 Sep 2001 11:45:06 -0700 (PDT)
From: Brian Wellington <Brian.Wellington@nominum.com>
To: bert hubert <ahu@ds9a.nl>
cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: should RCODE override a regular record in the zone?
In-Reply-To: <20010928104422.A3283@outpost.ds9a.nl>
Message-ID: <Pine.LNX.4.33.0109281144400.2389-100000@spratly.nominum.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

On Fri, 28 Sep 2001, bert hubert wrote:

> On Thu, Sep 27, 2001 at 07:57:26PM -0700, Brian Wellington wrote:
>
> > If, however, you are storing data in a database that stores the textual
> > representation of types, you should convert unknown RR type formatted
> > records into their canonical form before storing them, when the type is
> > known.  For example, if you see a TYPE1 record, store it as an A record.
>
> The main reason for being smart about 'known unknown types' is upgrades.
> Only then will you encounter a TYPE# which you now understand natively.

No.  The unknown rr-types draft does not prohibit using TYPE1 records
instead of A records, for example.

> > > In fact, it might be better to make each query an ANY query internally
> > > then.
> >
> > No, this would be an exceedingly bad idea, and is completely unnecessary.
>
> I suggest you write a database driven nameserver before making statements
> about them.

I suggest you don't make assumptions.  Having written both the unknown rr
type support for bind 9 and the sample postgres backend, I think I'm
qualified to say that it is possible, using exactly the method I
described.

> Anyway, we've implemented it because it is useful for us - it provides a lot
> of flexibility because of the ability to serve unknown types. I'm not sure
> if we'll follow the 'known unknown types are equivalent with natice types'
> rule.

Then you're violating the spec.

Brian



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 18:01:16 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA25822
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 18:01:14 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n5QT-000I0P-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 14:44:05 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n5QS-000I0J-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:44:04 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n5QH-0009fN-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:43:53 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <5.1.0.14.2.20010928141318.028528f0@localhost>
Date: Fri, 28 Sep 2001 14:17:29 -0400
To: Pekka Savola <pekkas@netcore.fi>, <namedroppers@ops.ietf.org>
From: =?iso-8859-1?Q?=D3lafur?= =?iso-8859-1?Q?_Gu=F0mundsson?=
  <ogud@ogud.com>
Subject: Re: I-D ACTION:draft-ietf-dnsext-ipv6-addresses-00.txt
In-Reply-To: <Pine.LNX.4.33.0109281643370.9190-100000@netcore.fi>
References: <200109281109.HAA09046@ietf.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

At 09:44 AM 9/28/2001, Pekka Savola wrote:
>On Fri, 28 Sep 2001 Internet-Drafts@ietf.org wrote:
> >       Title           : Representing IPv6 addresses in DNS
> >       Author(s)       : R. Bush, A. Durand, B. Fink, O. Gudmundsson, T. 
> Hain
> >       Filename        : draft-ietf-dnsext-ipv6-addresses-00.txt
> >       Pages           : 6
> >       Date            : 27-Sep-01
>
>Just a note; I support this but in all honesty, perhaps s/consensus/rough
>consensus/ would be closer to the community opinion. :-)


The point of the document is to judge consensus, editors tried to
capture the "sense of the room" in London and this is now put out
for judging what kind consensus is  for the recommendations, by the
working group.
Fire away, editors have put on their asbestos suits :-))
If there is no discussion, the co-chairs may start a last call real soon.

         Olafur



to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 18:02:02 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA25844
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 18:02:00 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n5S2-000I3F-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 14:45:42 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n5S2-000I39-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:45:42 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n5Rw-0009fW-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:45:36 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Fri, 28 Sep 2001 14:03:10 -0700 (PDT)
Message-Id: <200109282103.OAA05336@gulag.araneus.fi>
To: namedroppers@ops.ietf.org
Subject: draft-ietf-dnsext-ipv6-addresses-00.txt
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

draft-ietf-dnsext-ipv6-addresses-00.txt states:

    The hexadecimal text representation of IPv6 addresses appears to be
    capable of expressing all of the delegation schemes that we expect to
    be used in the DNS reverse tree, since we do not ever expect to see
    delegation in the least significant possible hexadecimal label.  That
    is, we do not ever expect to see an IPv6 address architecture that
    advocates address delegation in the least significant four bits of an
    IPv6 address.

As I tried to point out at the London meeting, there is absolutely
nothing special about the least significant label.  If someone wants
to delegate the reverse mapping of single IPv6 addresses to separate
zones each containing a single PTR record at its apex, they can.

The reason why RFC2317 makes the least significant octet a special
case when delegating IPv4 subnets is not that the general method of
multiple delegations doesn't work for the last octet -- it does.  It's
just that in the IPv4 case, delegations within the last octet are
common, and using the multiple delegation scheme for a /25 network
would mean that the leaf network owner would have to set up as many as
128 separate zones each containing a single PTR.  The RFC2317 CNAME
hack is merely a way to combine these into a single zone for the
convenience of the leaf network owner.

In IPv6, the corresponding /125 networks are so rare, and the
corresponding multiplicity of zones (8 rather than 128) so small,
that there is no need to make the last nibble a special case --
single-PTR zones will serve this purpose just fine.

Therefore, I suggest replacing the quoted paragraph by the following:

    Delegation of reverse mappings for networks whose prefix length is
    not a multiple of four bits can be expressed in the RFC1886 scheme
    by creating multiple delegations of consecutive hexadecimal
    labels.

-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Fri Sep 28 19:01:34 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA25756
	for <dnsext-archive@lists.ietf.org>; Fri, 28 Sep 2001 17:59:56 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15n5Rm-000I35-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 14:45:26 -0700
Received: from mpfg.attlabs.net ([12.106.35.2] helo=roam.psg.com)
	by psg.com with esmtp (Exim 3.33 #1)
	id 15n5Rm-000I2z-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:45:26 -0700
Received: from randy by roam.psg.com with local (Exim 3.33 #1)
	id 15n5Rg-0009fU-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 14:45:20 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: 28 Sep 2001 19:44:54 -0000
Message-ID: <20010928194454.10555.qmail@cr.yp.to>
From: "D. J. Bernstein" <djb@cr.yp.to>
To: namedroppers@ops.ietf.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <20010927052308.691.qmail@cr.yp.to> <200109271613.f8RGDar96471@drugs.dv.isc.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Mark.Andrews@isc.org writes:
> The existing RFC's failed to cover master file formats for
> unknown types.  As such they were incomplete.

This issue is explicitly addressed in RFC 1123:

   A name server may acquire, via zone transfer, RRs that the server
   doesn't know how to convert to printable format. ... For proper
   operation, this data must be preserved, and hence the implication is
   that DNS software cannot use textual formats for internal storage.

It's not my fault that you haven't read RFC 1123. The tinydns data
format has supported unknown types for nearly two years.

I understand that you've finally extended the BIND input format to allow
unknown types. I have no objections to documenting that format. I do
object, however, to your pretense that BIND's previous behavior was
acceptable.

---Dan


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sat Sep 29 01:32:13 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA04134
	for <dnsext-archive@lists.ietf.org>; Sat, 29 Sep 2001 01:32:13 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15nCIS-0001gt-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 22:04:16 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15nCIS-0001gn-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:04:16 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15nCIS-000EZE-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:04:16 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: 29 Sep 2001 00:51:36 -0000
Message-ID: <20010929005136.19379.qmail@cr.yp.to>
From: "D. J. Bernstein" <djb@cr.yp.to>
To: namedroppers@ops.ietf.org
Subject: Re: Minimize security-critical code
References: <20010927174557.30969.qmail@cr.yp.to> <200109280159.f8S1xtr99782@drugs.dv.isc.org>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Mark.Andrews@isc.org writes:
> The bug exists in more versions
  [ ... ]
> but it cannot be triggered in the earlier versions

You're splitting hairs. The bottom line is that CERT Advisory 2000-20
would never have happened if you had treated all new record types as
opaque data. Other implementors should be encouraged to _avoid_ your
mistakes, not to imitate them.

---Dan


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sat Sep 29 01:32:17 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA04149
	for <dnsext-archive@lists.ietf.org>; Sat, 29 Sep 2001 01:32:16 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15nCIz-0001iG-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 22:04:49 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15nCIy-0001iA-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:04:48 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15nCIy-000EZb-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:04:48 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109290112.f8T1Co904901@drugs.dv.isc.org>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
From: Mark.Andrews@isc.org
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt 
In-reply-to: Your message of "28 Sep 2001 19:44:54 GMT."
             <20010928194454.10555.qmail@cr.yp.to> 
Date: Sat, 29 Sep 2001 11:12:50 +1000
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Mark.Andrews@isc.org writes:
> > The existing RFC's failed to cover master file formats for
> > unknown types.  As such they were incomplete.
> 
> This issue is explicitly addressed in RFC 1123:
> 
>    A name server may acquire, via zone transfer, RRs that the server
>    doesn't know how to convert to printable format. ... For proper
>    operation, this data must be preserved, and hence the implication is
>    that DNS software cannot use textual formats for internal storage.

	Which just proves to point that RFC 1034 / RFC 1035 were
	incomplete as they didn't meet their intent of allowing
	zones to be moved between nameservers (from different vendors)
	using textual means.

	The correct fix when this was identified back when 1123 was
	written should have been to do what we are doing now and
	define a mechanism.

	Mark
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sat Sep 29 01:32:21 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA04164
	for <dnsext-archive@lists.ietf.org>; Sat, 29 Sep 2001 01:32:20 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15nCD4-0001aq-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 21:58:42 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15nCD4-0001ak-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 21:58:42 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15nCD3-000EWk-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 21:58:42 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Fri, 28 Sep 2001 16:14:09 -0700 (PDT)
Message-Id: <200109282314.QAA05401@gulag.araneus.fi>
To: Kevin Darcy <kcd@daimlerchrysler.com>
Cc: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
In-Reply-To: <3BB4E766.11981711@daimlerchrysler.com>
References: <E15m6zI-000NCA-00@rip.psg.com>
	<20010926111745.A10616@outpost.ds9a.nl>
	<200109261904.MAA03755@gulag.araneus.fi>
	<3BB23A32.4FB62E0D@daimlerchrysler.com>
	<200109262147.OAA03864@gulag.araneus.fi>
	<3BB251E7.E77BE8AB@daimlerchrysler.com>
	<200109271742.KAA04466@gulag.araneus.fi>
	<3BB3A0A7.3B63FB1B@daimlerchrysler.com>
	<200109281744.KAA05226@gulag.araneus.fi>
	<3BB4E766.11981711@daimlerchrysler.com>
From: gson@isc.org (Andreas Gustafsson)
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

Kevin Darcy writes:
> If all you're doing is "documenting" requirements, shouldn't you at least
> be noting the fact that the prohibition -- or at least the rationale for
> the prohibition -- is partially obsoleted by EDNS0?

EDNS0 does not change the rules for compression - it's a general
purpose negotiation mechanism that could be used to change any number
of aspects of the DNS protocol, including the rules for compression.
If some future EDNSn (n>0) does this, it will obsolete certain parts
of the unknown-rrs specification, but EDNS0 itself does not.

> Then turn the extension OFF if you're behind one of those beasts!

Then you are imposing the requirement that extended compression is
configurable, and the administrative burden of manually configuring
it.
-- 
Andreas Gustafsson, gson@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sat Sep 29 01:33:22 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA04255
	for <dnsext-archive@lists.ietf.org>; Sat, 29 Sep 2001 01:33:21 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15nCNv-0001qY-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 22:09:55 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15nCNv-0001qS-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:09:55 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15nCNv-000Ebt-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:09:55 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: namedroppers@ops.ietf.org
Subject: Re: I-D ACTION:draft-ietf-dnsext-ipv6-addresses-00.txt
References: <200109281109.HAA09046@ietf.org> <5.1.0.14.2.20010928141318.028528f0@localhost>
From: Paul Vixie <vixie@vix.com>
Date: 28 Sep 2001 19:39:17 -0700
In-Reply-To: Ólafur  Guðmundsson's message of "28 Sep 2001 14:53:32 -0700"
Message-ID: <g3ofnud8kq.fsf@as.vix.com>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> The point of the document is to judge consensus, editors tried to
> capture the "sense of the room" in London and this is now put out
> for judging what kind consensus is  for the recommendations, by the
> working group.  Fire away, editors have put on their asbestos suits :-))
> If there is no discussion, the co-chairs may start a last call real soon.

this i-d kills ipv6 for any practical purpose beyond 3G deployment, by
raising the renumbering penalty beyond any possible benefit, and by
moving the multihoming problem back into the routing system where there
is no current hope of solving it.

this i-d is an implicit declaration that ipv4 with NAT is the future of
the internet and that we had all better stop spending time on ipv6 other
than to make sure wireless clouds and other ieee MAC mappings can use it
at the edges.


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sat Sep 29 01:34:15 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id BAA04327
	for <dnsext-archive@lists.ietf.org>; Sat, 29 Sep 2001 01:34:14 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15nCKj-0001lV-00
	for namedroppers-data@psg.com; Fri, 28 Sep 2001 22:06:37 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15nCKi-0001lO-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:06:36 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15nCKi-000EaU-00
	for namedroppers@ops.ietf.org; Fri, 28 Sep 2001 22:06:36 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <3BB52A7D.5C859040@daimlerchrysler.com>
Date: Fri, 28 Sep 2001 21:57:17 -0400
From: Kevin Darcy <kcd@daimlerchrysler.com>
To: namedroppers <namedroppers@ops.ietf.org>
Subject: Re: wg last call for draft-ietf-dnsext-unknown-rrs-01.txt
References: <E15m6zI-000NCA-00@rip.psg.com>
		<20010926111745.A10616@outpost.ds9a.nl>
		<200109261904.MAA03755@gulag.araneus.fi>
		<3BB23A32.4FB62E0D@daimlerchrysler.com>
		<200109262147.OAA03864@gulag.araneus.fi>
		<3BB251E7.E77BE8AB@daimlerchrysler.com>
		<200109271742.KAA04466@gulag.araneus.fi>
		<3BB3A0A7.3B63FB1B@daimlerchrysler.com>
		<200109281744.KAA05226@gulag.araneus.fi>
		<3BB4E766.11981711@daimlerchrysler.com> <200109282314.QAA05401@gulag.araneus.fi>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

gson@isc.org wrote:

> Kevin Darcy writes:
> > If all you're doing is "documenting" requirements, shouldn't you at least
> > be noting the fact that the prohibition -- or at least the rationale for
> > the prohibition -- is partially obsoleted by EDNS0?
>
> EDNS0 does not change the rules for compression - it's a general
> purpose negotiation mechanism that could be used to change any number
> of aspects of the DNS protocol, including the rules for compression.
> If some future EDNSn (n>0) does this, it will obsolete certain parts
> of the unknown-rrs specification, but EDNS0 itself does not.

EDNS0 undercuts the rationale for prohibiting compression in new record types,
because it provides a "safe", point-to-point way to compress records of those
types. Agreed, that's not technically the same as "obsoleting" the
prohibition, but it does raise a significant likelihood that exceptions will
become necessary. I think there should at least be some language in the
document acknowledging this fact, because -- let's face it -- you're not just
"documenting" an existing prohibition, you're *reinforcing* it.

> > Then turn the extension OFF if you're behind one of those beasts!
>
> Then you are imposing the requirement that extended compression is
> configurable, and the administrative burden of manually configuring
> it.

It wouldn't necessarily be a manual operation, and it wouldn't necessarily be
burdensome. If the new record type has only a specialized use, then perhaps
implementations will default it OFF and only those administrators who go out
of their way to do so -- presumably with a full understanding of their local
network/DNS environment and/or the tradeoffs of using the new record type --
will turn it ON.


- Kevin




to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


From owner-namedroppers@ops.ietf.org  Sat Sep 29 12:18:10 2001
Received: from psg.com (exim@psg.com [147.28.0.62])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA23055
	for <dnsext-archive@lists.ietf.org>; Sat, 29 Sep 2001 12:18:10 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.33 #1)
	id 15nMQs-000Cfp-00
	for namedroppers-data@psg.com; Sat, 29 Sep 2001 08:53:38 -0700
Received: from rip.psg.com ([147.28.0.39])
	by psg.com with esmtp (Exim 3.33 #1)
	id 15nMQr-000Cfj-00
	for namedroppers@ops.ietf.org; Sat, 29 Sep 2001 08:53:37 -0700
Received: from randy by rip.psg.com with local (Exim 3.33 #1)
	id 15nMQr-0006jZ-00
	for namedroppers@ops.ietf.org; Sat, 29 Sep 2001 08:53:37 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <200109290715.f8T7FZ905817@drugs.dv.isc.org>
To: "D. J. Bernstein" <djb@cr.yp.to>
Cc: namedroppers@ops.ietf.org
From: Mark.Andrews@isc.org
Subject: Re: Minimize security-critical code 
In-reply-to: Your message of "29 Sep 2001 00:51:36 GMT."
             <20010929005136.19379.qmail@cr.yp.to> 
Date: Sat, 29 Sep 2001 17:15:35 +1000
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

> Mark.Andrews@isc.org writes:
> > The bug exists in more versions
>   [ ... ]
> > but it cannot be triggered in the earlier versions
> 
> You're splitting hairs. The bottom line is that CERT Advisory 2000-20
> would never have happened if you had treated all new record types as
> opaque data. Other implementors should be encouraged to _avoid_ your
> mistakes, not to imitate them.
> 
> ---Dan

	Sure they should be encouraged not to make the same mistake.

	However you need to know what the mistake is before you
	start encouraging people not to do something.  You quite
	evidently don't know what the actual mistake was or you
	would not be complaining about SRV records being treated
	as non-opaque data.  You would/should be encouring them to
	fix the root cause of the problem rather than spouting
	nonsense about ignoring the rdata field contents of known
	RR types.

	Mark
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org


to unsubscribe send a message to namedroppers-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.


