[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


To: "Jaehoon Paul Jeong" <paul@etri.re.kr>
Cc: <dnsop@cafax.se>
From: John Schnizlein <jschnizl@cisco.com>
Date: Mon, 04 Aug 2003 07:22:13 -0400
In-Reply-To: <009d01c35a4a$32a8f840$c470fe81@etri.re.kr>
Sender: owner-dnsop@cafax.se
Subject: Re: Policy of IPv6 DNS Discovery

At 01:35 AM 8/4/2003, Jaehoon Paul Jeong wrote:

>Where can we discuss new protocol extensions if neither in IPv6 wg, nor in DNSOP wg?
>This is a question for all DNSOP members except for you.

The discussion of the question whether a protocol extension is
necessary for IPv6 hosts to locate DNS servers is underway here.
All the chair is requesting is that we understand first if there
is any need, before we discuss the nature of protocol extensions
that might fit that need.

>I think, I did my best to appeal the necessity of RA-based DNS Discovery, 
>although I could not express my opinions well enough because of the limitation of Non-English speaker.

Your interest is clear, and those of us who understand only English
appreciate your extra effort in making it clear in a language other
than your first.

>If there is proper reason RA-based DNS Discovery SHOULD NOT coexist 
>with DHCPv6, I will also rescind my concerns about RA-based DNS Discovery.

I think the reason is stated in RFC 1958: Architectural Principles of 
the Internet.:

   3.2 If there are several ways of doing the same thing, choose one.
   If a previous design, in the Internet context or elsewhere, has
   successfully solved the same problem, choose the same solution unless
   there is a good technical reason not to.  Duplication of the same
   protocol functionality should be avoided as far as possible, without
   of course using this argument to reject improvements.

While this does not deny that different solutions might be appropriate
for different operational situations, it does require that the nature of
the different situations, and the suitability of proposed solutions to 
those differences, should be identified. We have not seen the difference.

Since there is a protocol (DCHPv6) which can provide the necessary
information for hosts, and since the "heavy-weight" concerns attributed
to that protocol have been shown invalid (existence proof: DHPCv6-light),
justification for developing a new protocol (or extension of RA) must meet 
the burden identified in Architectural Principle 3.2

John

#----------------------------------------------------------------------
# To unsubscribe, send a message to <dnsop-request@cafax.se>.

Home | Date list | Subject list