Multicast DNS clarification needed

"Nicolae, Dan" <Dan.Nicolae@usa.xerox.com> Wed, 16 August 2000 00:08 UTC

Received: from psg.com (psg.com [147.28.0.62]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id UAA00969 for <dnsext-archive@lists.ietf.org>; Tue, 15 Aug 2000 20:08:07 -0400 (EDT)
Received: from lserv by psg.com with local (Exim 3.13 #1) id 13Opva-000IAN-00 for namedroppers-data@psg.com; Tue, 15 Aug 2000 16:15:26 -0700
Received: from rip.psg.com ([147.28.0.39]) by psg.com with esmtp (Exim 3.13 #1) id 13OpvZ-000IAH-00 for namedroppers@ops.ietf.org; Tue, 15 Aug 2000 16:15:25 -0700
Received: from randy by rip.psg.com with local (Exim 3.13 #1) id 13OpvZ-000A5y-00 for namedroppers@ops.ietf.org; Tue, 15 Aug 2000 16:15:25 -0700
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Date: Tue, 15 Aug 2000 19:12:46 -0400
From: "Nicolae, Dan" <Dan.Nicolae@usa.xerox.com>
Subject: Multicast DNS clarification needed
To: namedroppers@ops.ietf.org
Message-id: <3654E69400ADD211A3A400805FA7CE24022A6300@USA0111MS2>
Sender: owner-namedroppers@ops.ietf.org
Precedence: bulk
Content-Transfer-Encoding: 7bit

First of all, before commenting over the MDNS drafts draft it would be good
to know if draft-aboba-dnsext-mdns-01.txt and Bill Manning's draft, both
with similar names, are being seen as competing specifications or they are
supposed to coexist.

I am assuming here that, despite the identical name, the two drafts
complement each other and commenting on one involves keeping the other in
mind.


I remember seeing remark about the practical impossibility to have two
separate processes listening on the same IP port but different IP addresses
(unicast and multicast). I am not sure how true that is, but if it is, it is
important to consider it when talking about three mechanisms (DNS and two
MDNS) that plan to share the same port (53).

This issue was mentioned in the SSDP draft (now defunct). Here is the
relevant paragraph:
 " 8.3.2.    Why does SSDP need to use a port other than 80? 
 
   The result is that if we used port 80 on the SSDP multicast scope 
   then we would require that the SSDP software also grab port 80 for 
   the local machine. This would mean that SSDP could only be 
   implemented on machines which either didn't have HTTP servers or 
   whose HTTP servers had been enhanced to support SSDP. 
    
   We felt this was a unnecessary restriction. Therefore we are 
   choosing to use a port other than 80 on the SSDP multicast channel. "

Assuming that all three DNS protocols may have to be implemented in a single
process, and the DNS server behavior is dependent on what IP address a DNS
packet is received from (unicast or the two multicast groups), I see
potential implementation problems.

I hope some sockets gurus will pick this one up and validate it for various
flavors of OSes and sockets.

Then we'll know if the two drafts can be treated separately or all three
(add standard DNS) must be judged together from a single process
perspective.

Thanks,
/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.