On Thu, 27 Aug 1998, TTSG wrote:
> Date: Thu, 27 Aug 1998 19:42:50 -0400 (EDT)
> From: TTSG <ttsg@ttsg.com>
> To: Russell Van Tassell <russell@cscorp.com>
> Cc: vikas@navya.com,
jonz@cartman.netrail.net,
nocol-users@navya.com
> Subject: Re: nsmon question
>
> > Well, me, I have about a half dozen name servers that I monitor, first
> > of all, just to make sure named remains available... so, in that sense,
> > it makes no sense for me to see a half dozen identical domain names in
> > netconsole and still not know where it's failed.
> >
> > So, my opinion would be that IF there's a AUTH statement on the line,
> > perhaps it should be "domainname-host" (or vice-versa) that comes up in
> > netconsole rather than just the domainname. Without the AUTH parameter,
> > it should probably just be the host name (though I haven't used such a
> > statement in 4.2... yet)
> >
> I don't quite agree. I monitor over 60 domains, strewn
> over about 2 dozen machines. Some of the names of the domains are long
> (15 characters) and some of the machines names are long (hostname is
> 8). This is pretty long if you were to do that. I think I'd rather
> come up with my own little codes.
Yeah, I have this same "problem" at another site... some domain names
are longer than that (we're talking names that, I think, are abusing the
long length "features" of DNS); and we're talking in numbers a few
orders magnitude greater... so that's where a good web interface comes
in...
> Without AUTH, why still wouldn't you want to know what name of
> the machine that isn't able to come up with the answer. The root name
> servers aren't authoritive, but I'd still like to know which one THIS
> time claims my domain doesn't exist.
>
> Tuc/TTSG
That was sort of my point, too... but, at the same rate, I'd like to
know (at a glance) which name server was having problems AND what domain
it was foobar'ing on...
Russell
--
Russell M. Van Tassell Commercial Systems Corp.
russell@cscorp.com http://www.cscorp.com/
Ph: 1-888/COMM-SYS
|