IETF Logo Mail Archive

Re: Should a nameserver know about itself?

Sam Trenholme <namedroppers@artemas.reachin.com> Thu, 10 May 2001 20:47 UTC

Received: from nic.cafax.se ([192.71.228.17]) by ietf.org (8.9.1a/8.9.1a) with SMTP id QAA08008 for <dnsop-archive@odin.ietf.org>; Thu, 10 May 2001 16:47:27 -0400 (EDT)
Received: by nic.cafax.se (8.12.0.Beta5/8.12.0.Beta5) id f4AKNUpP010837 for dnsop-outgoing; Thu, 10 May 2001 22:23:30 +0200 (MEST)
Received: from naptop.autonomica.se (flaptop.liman.sunet.se [193.10.90.102]) by nic.cafax.se (8.12.0.Beta7/8.12.0.Beta5) with ESMTP id f4AKNQLt010832 for <dnsop@cafax.se>; Thu, 10 May 2001 22:23:26 +0200 (MEST)
Received: by naptop.autonomica.se (8.12.0.Beta1/8.12.0.Beta1) id f4AKLtQu000483 for dnsop@cafax.se; Thu, 10 May 2001 22:21:55 +0200 (MEST)
Received: from artemas.reachin.com (artemas.reachin.com [64.14.214.33]) by nic.cafax.se (8.12.0.Beta7/8.12.0.Beta5) with SMTP id f4A8fZLt006435 for <dnsop@cafax.se>; Thu, 10 May 2001 10:41:36 +0200 (MEST)
Received: (qmail 31937 invoked by uid 1233); 10 May 2001 01:41:34 -0700
Received: from localhost (sendmail-bs@127.0.0.1) by localhost with SMTP; 10 May 2001 01:41:34 -0700
Date: Thu, 10 May 2001 01:41:34 -0700
From: Sam Trenholme <namedroppers@artemas.reachin.com>
To: Jim Reid <Jim.Reid@nominum.com>
cc: Mark.Andrews@nominum.com, Bruce Campbell <bruce.campbell@apnic.net>, dnsop@cafax.se
Subject: Re: Should a nameserver know about itself?
In-Reply-To: <30735.989481354@shell.nominum.com>
Message-ID: <Pine.LNX.4.30.0105100136270.31920-100000@artemas.reachin.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset="US-ASCII"
Sender: owner-dnsop@cafax.se
Precedence: bulk

>     Mark> 	Send it a non-recursive query.  A nameserver should
>     Mark> always answer even if it is just a header containing
>     Mark> refused, servfail or notimp.
>
> Yes, but there is one notorious DNS implementation that doesn't do
> that. It fails to return any answer -- not even a referral for . -- if
> it's asked for a name that it isn't authoritative for.

Fortunatly, his other suggestion about sending a status opcode will get a
"not implemented" response from the "nortorious implementation" in
question.

The only DNS server I know about that has version out there that don't
correctly handle the "status" opcode is my own implentation of a DNS
server.  It used to handle a status query like a normal query.  In fact,
Mark's posting made me realize that my implementation had this problem.
In light of his posting, I just released a new bugfix release because of
this issue.

- Sam

v2.23.0 | Report a Bug | By Email