To:
rad@twig.com (Richard Doty)
Cc:
dnsop@cafax.se
From:
Johan Ihren <johani@autonomica.se>
Date:
02 Feb 2002 13:04:17 +0100
In-Reply-To:
rad@twig.com's message of "Thu, 31 Jan 2002 20:42:07 +0000"
Sender:
owner-dnsop@cafax.se
User-Agent:
Gnus/5.070095 (Pterodactyl Gnus v0.95) Emacs/20.3
Subject:
Re: draft-ietf-dnsop-v6-name-space-fragmentation-00.txt
rad@twig.com (Richard Doty) writes: Hi, > I do not understand the reason for using only single-stack nameservers: > > 6.2. How to deploy DNS data. > ... > It is recommended that the name servers run on single stack > machines, i.e. machines that are only able to utilize either IPv4 > transport or IPv6 transport, but not both. That paragraph will go. While there are benefits (simplicity being one) to avoiding dual stack nameservers this document is not the place to argue that case. > Also, some of the suggestions, e.g. > > 4.1. Requirement of IPv4 address for at least one name server. > > seem to beg a reference to rfc 2182. By recommending "at least one" > name server reachable from a particular version of IP, the effect for > hosts that use that version of IP exclusively is that zones are > configured with only one nameserver. I am the first one to agree to the downsides to only having one nameserver for a zone. However, I don't see this document as a replacement for 2182 in any way, i.e. the guidelines in 2182 still apply. The point with my document is to argue the case that a zone that does not really care about v4 to still maintain presence in v4, thereby avoiding fragmentation of the namespace (as seen from v4). I'll try to re-word this without getting bogged down in actual numbers, referring that issue to better suited documents. Thanks, Johan