The Samba-Bugzilla – Bug 10108
Winbind takes way too long to fail to cached logins if all AD servers down
Last modified: 2014-01-29 00:07:01 UTC
In the worst case situation, if you lose ALL of your AD servers, winbind takes a very, very long time to fail to iusing cached logins instead.
I see of no way of forcing winbind to abort trying to contact an AD server if they're all down, that results in an unacceptable wait to log into a server.
Could we please have an option to let us set a timeout for when winbind will give up trying to see if a domain controller is available or if they're all down fail to local cached logins quicker somehow?
can you provide a network sniff and a rough overview of the setup that you have in the case that you describe?