On Sun, 1 Apr 2001, Robin Szemeti wrote:
> On Sun, 01 Apr 2001, you wrote:
> > On Sun, 1 Apr 2001, Robin Szemeti wrote:
> > > agreed it is a dumb thing, especially if your nameserver doesnt have a
> > > name to lookup ....
> > Erm!!?!? How exactly were you planning to point anything at it? an NS
> > RR requires an authoritative name as it's RHS.
> if its a box-over-in-the corner that one day will be your DNS server
> somewhere but right now its just a ip address on a network you're trying
> to test before deploying .. it did get a name eventually.
Hmmm.. I don't quite know how you can *test* it, if it hasn't got any
names, and therefore can't serve any zones...
> and .. surely not all the nameservers are necessarily named, only the ones
> published to the world, you could have internal servers that don;t answer
> external queries (such as a primary master server with two slaves used as
> authoratative servers to the world, whilst your primary master is left
> untroubled) in which case the primary master would not need a name just
> an address, and you might want to query it directly yourself to make sure
> it was not telling porkies. ??
*burble*! It is reasonable to have this property, yes, but then what do
you put in the host part of the SOA record for the zones served by this.
The entry there should be the master server for the zone. The other case
that you might want this is for a resolver, however again, the small
amount of work involved in assigning a name for the machine in question
suggests that you get no added benefit by *not* doing so.
MBM
--
Matthew Byng-Maddick Home: <[EMAIL PROTECTED]> +44 20 8980 5714 (Home)
http://colondot.net/ Work: <[EMAIL PROTECTED]> +44 7956 613942 (Mobile)
Tell me, O Octopus, I begs, / Is those things arms, or is they legs? /
I marvel at thee, Octopus; / If I were thou, I'd call me us. -- Ogden Nash
