I know this is a Solaris 9 forum, but I couldn't find a Solaris 8.
Basically, I'm having problems with winbindd crashing. While I'm watching ldap requests with snoop I see about 90% good traffic, then these stubborn IP's of domain controllers of a different domain on another network. I don't know how they got there, but too many of them and winbindd crashes with a PANIC in the log file. I do have a 2 way non transitive trust with this domain. I know this isn't a lot of info but maybe it's as simple as clearing out some sort of cache somewhere.
I had a bizarre problem with NIS+ many years ago that might be related to your issue. My NIS+ master server was connected to 2 networks, one for normal WAN traffic and one restricted to only backups (local LAN only connection). On a periodic basis, I think every hour, the NIS+ master communicates wiht its slaves (or whatever the politically correct term is now). For some reason, this periodic update failed, which resulted in the entire system going down. It took multiple unix, network, and SUN support staff 12 hours to figure out that the NIS+ master server was attempting to communicate to the slaves only through the backup LAN network and not through the WAN network. The LAN was on bge0 and the WAN was on bge1. The fix was to put the WAN on bge0 and the LAN on bge1.
Any chance that the netmask is incorrect?
Edited by: Paisley on Feb 18, 2011 3:49 PM