To:
dnssec@cafax.se
From:
Miek Gieben <miekg@atoom.net>
Date:
Thu, 13 May 2004 14:57:43 +0200
Content-Disposition:
inline
In-Reply-To:
<20040510215727.83AB542B2@thrintun.hactrn.net>
Mail-Followup-To:
dnssec@cafax.se
Sender:
owner-dnssec@cafax.se
User-Agent:
Vim/Mutt/Linux
Subject:
Re: dnssec: resolver - application communication
[On 10 May, @23:57, Rob wrote in "Re: dnssec: resolver - applica ..."] > please see the handling of the cd bit as specified in the bis drafts, > and if -that- is wrong, please tell us asap. I think the definition is correct. I also believe (as stated before) that SERVFAIL can be used as a NOTVAL: suppose the following happens, a resolver talks to a cache: SERVFAIL -> requery with CD=1 -> SERVFAIL -> "real" SERVFAIL SERVFAIL -> requery with CD=1 -> answer -> "NOTVAL" So: receiving two SERVFAILS signals something really broken, and receiving one SERVFAIL signal a NOTVAL. I don't see why: NOTVAL -> requery with CD=1 -> answer is any better (expect if one wants to avoid a few unneeded requeries), grtz, Miek