Domains not resolving in Bind9, flush cache, clears issue.

[Brook Harty]

Jim Reid <jim at rfc1035.com> wrote in message news:<c5hht1$lbd$1 at sf1.isc.org>...
> >>>>> "Brook" == Brook Harty <harty at ironwolve.com> writes:
> 
>     Brook> Yup, now I see all the domains having issues are glue
>     Brook> record issues.  Must be a common problem when upgrading to
>     Brook> Bind9.  How do larger ISP's avoid this problem, just stick
>     Brook> with bind8?
> 
> No. They don't get themselves into situations where they rely on
> broken glue setups and suchlike that legacy DNS software tolerated. If
> your zone files and delegations are in good health -- ie as they're
> supposed to be -- using BIND9 presents no problem at all.
> 
> Your problems are caused by self-inflicted wounds. icee is delegated
> to mailscan.pretzelgourmet.com and mailscan.gourmettwists.net. These
> two zones are delegated to these two servers by the .com and net
> zones. So far, so good. These two name servers are authoritative for
> the pretzelgourmet.com and gourmettwists.net zones. But both zones --
> the definitive source of information about them! -- say that
> mailscan.pretzelgourmet.com and mailscan.gourmettwists.net don't
> exist! This is why your zones are unresolvable. Add A records for
> these names to the zones, just like you were supposed to have done,
> and your problem will go away.

I think you missed my post, they are not my domains, they are external
domains internal customers are looking up.  What I'm wondering is how
you bypass the problem on these external domains, that are setup
incorrectly. I use Bind8 my customers can resolve these incorrectly
setup domains, if I use Bind9 they will time out when the glue
expires.  Did I step my foot into a common issue?

As for the post about lame log, I turn the logging off due to the size
of the log files just for lame errors, its amazing how many lame
servers are out there.