Hmmmm....
If you catch the issue, before correcting it, please turn on addebug and have the user try again and fail (this will give us a good repro).
You can try to look at /var/log/messages for Centrify Audit Trail events in the minutes prior to the failure this morning, this may contain some information that is relevant.
Unfortunately, sporadic issues are hard to troubleshoot because it is hard to get a repro.
It's also important to understand the output of adinfo -y netstate and if the system is trying to talk to an unreachable DC. If the AD topology contains unreachable DCs (quite common when DMZ or RODCs aren't blacklisted).
If you catch a log, please tag us on this thread and we'll look at your logs.