[Condor-users] need to restart condor to restart idle jobs

["Upton, Stephen C" <scupton@xxxxxxx>]

Title: need to restart condor to restart idle jobs

Hi All,

The subject understates my issue; actually, I'm not sure what my problem is. ;-(

I have a master node that's running condor 7.2 on INTEL/OSX. My compute nodes are INTEL/WINNT 5.1.

1. I remotely log in as user condor on the OSX node, and get a "condor_status", all my machines show up.
2. I submit some WINNT 5.1 jobs, the jobs start fine.
3. I log out.  
4. I log back in to check the status, and the jobs have gone idle.
5. I do a "condor_restart"
6. I then do  a "condor_master" since the condor_restart doesn't seem to have restarted the master. Probably one (THE?) issue?
7. The jobs start running after about 5 minutes -probably some configuration parameter I haven't figured out yet. ;-(
8. I log out.
9. I log back in and submit some WINNT 5.1 jobs, and the jobs stay idle.
   

There's probably yet another Linux/OSX security/permissions dohickey I haven't figured out. I hope the above scenario is sufficient to suggest places to look, configuration parameters I need to adjust, etc.. I should also mention that the OSX host recently replaced the prior Linux host, in which everything was working, so I'm guessing that install is different in some way.

Any help would be greatly appreciated.

thanx!
steve

Stephen C. Upton
Research Associate
SEED (Simulation Experiments & Efficient Designs) Center for Data Farming
Naval Postgraduate School
Cell: 831-402-3888