When Domain Resolution just Stops one Day.

[Martin McCormick]

I may have found the reason, but here is the relevant
information. Our zone is okstate.edu and the zone that suddenly
broke is osu-com.okstate.edu which has only been there since
1998 as my Email archives turned up. It simply has not changed
in its configuration since we set it up for that campus in
either late 1997 or early 1998.

	What happened last week was that the address for
www.healthsciences.okstate.edu spontaneously stopped resolving
via a system on our campus that is a slave to the okstate.edu
zone but didn't know about osu-com.okstate.edu until yesterday
after at least 4 years of flawless operation.

	Querying ns.cis.okstate.edu is no problem. It's
essentially caching-only DNS's that sometimes loose these slave
deligated zones.

	The other time this happened to us, I had set up some
Microsoft Active Directory zones as subdomains to okstate.edu.
Our Oklahoma City campus has a local DNS running the same
version of bind we have here which is now 9.3.4. It had the
okstate.edu zone slaved but had none of the AD zones also slaved
because I thought it would just do the lookup against
okstate.edu.

	It appeared to do just that for a couple of years and
then that ringing phone again.

	I slaved all the AD zones to the Oklahoma City box and,
voila, no further problems.

	My real question is, why didn't it just fail to work
from the beginning or, more positively, how can I make sure
these deligated zones don't run for 3 or 4 years and then quit
for no good reason. Actually, there is a good reason, I just
haven't figured out what I am doing wrong on the slave
configurations.