bind not resolving localy

[Simon Waters]

Kalin Mintchev wrote:
> 
> hi list,
> 
> i have a redhat 7.2 box with bind 9.1.3 on it.
> bind is running. the first nameserver on the machine is the same IP as the
> machine itself. when i do dig for my.daomain.com the query defaults to the
> second name server. if i force it to look on the same machine with dig
> @this.machine i get the right results.

Why not 9.2.1?

Does you name server allow recursion?

> the logs don't log anything weird
> i can't find anything wrong with the named.conf..

You said it answers queries when asked so maybe nothing is
wrong.

> this wouldn't be a problem except it slows down the ftp, ssh and other
> sessions...

You said a second name server was available, so nothing should
slow down for as long as your back up nameserver isn't on the
other side of the planet.

I'm thinking perhaps your name servers have different views of
the DNS, if so you need to rethink the design.

> also if i do ping to something.whatever.com i get huge dropouts, but if i
> ping the same machine using the IP address the ping is fine...

Since ping first resolves, and then pings, assuming it reports
the same address, this should make no difference to packet loss.
Conceivable the network is congested with broken DNS queries or
some such, but I suspect your missing something more basic.