Comment 15 for bug 667269

I just did a backtrace on another machine(bee37), of another core file, the back trace takes a different path, but the final 3 calls are the same as my previous post and the same last e calls as Mau's back trace:

rpc_pipe_np_smb_conn
rpccli_set_timeout
winbindd_lookup_sids
...

Unfortunately these 3 back traces that have the same last 3 calls, are all different than my original backtrace I posted on 12/27/2010,
but I'll take any winbind fix for any problem at this point.

I had 22 of 80 machines experience a winbind
problem yesterday, Monday 2011-04-18.

I will throw one red herring of an idea out there.
Is it possible that the cause of this is
if a user is in a Windows group per his poilicy file but that group isn't active anymore?
winbind can work for days at a time, so this problem is not deterministic. There's some
other factor that is causing this to happen, and I can't figure it out.