This sounds a lot like bug 16355, which we ran into earlier this year. The filer incorrectly registers itself with WINS using a workstation(00), messenger (03), and server (20) that matches the domain name (1c). IE, there's the correct entries for a machine named FILER with 00, 03, and 20; then there's additional entries of DOMAIN with 00, 03, and 20. NetApp's support didn't have a solution, but someone on the list here did. NetApp support later confirmed that yes, it was bug number 16355, and the solution was fixed by upgrading to a newer version on OnTap.
-warren
Warren M. Walsh Honeywell Laboratories
-----Original Message----- From: Tom "Mad Dog" Yergeau [mailto:MadDog@fool.com] Sent: Monday, December 11, 2000 1:18 PM To: 'toasters@mathworks.com' Subject: cifs testdc problem and 1D registration conflict
We'd been seeing a problem where a cifs testdc would yield the following results:
filer> cifs testdc Using Established configuration Current Mode of NBT is H Mode
Netbios scope "" Registered names... FILER < 0> WINS FILER < 3> WINS FILER <20> WINS DOMAIN < 0> WINS
Testing Primary Domain Controller found 1 addresses trying 192.168.0.10...192.168.0.10 is alive found PDC DC-001
Testing all Domain Controllers found 3 addresses trying 192.168.0.10...192.168.0.10 is alive trying 192.168.0.10...192.168.0.10 is alive trying 192.168.0.13...192.168.0.13 is alive found DC DC-001 at 192.168.0.10 Not able to communicate with DC 192.168.0.10 found DC DC-002 at 192.168.0.13
NetApp was perplexed, had me re-run CIFS setup, etc. to no avail.
Today we rebooted our PDC because we noticed that the master browser registration was in conflict:
NetBIOS Remote Machine Name Table
Name Type Status --------------------------------------------- PDC <00> UNIQUE Registered PDC <20> UNIQUE Registered DOMAIN <00> GROUP Registered DOMAIN <1C> GROUP Registered DOMAIN <1E> GROUP Registered PDC <03> UNIQUE Registered DOMAIN <1D> UNIQUE Conflict ..__MSBROWSE__.<01> GROUP Registered DOMAIN <1B> UNIQUE Registered
After the PDC rebooted, the cifs testdc problem went away.
I know I've corresponded with some folks from toasters who were having this cifs testdc problem. If you're having the problem, can you do an nbtstat -a on your PDC, and see if you see a <1D> entry in conflict?
Thanks, Mad Dog
For your enjoyment: http://support.microsoft.com/support/kb/articles/Q163/4/09.asp