I have a condor pool composed of the central manager (running Windows) that is also the submitting machine but not an execute machine.
The pool has two execute nodes (Linux operative system).
When I submit a job, the state of the execute nodes changes from Unclaimed to Matched and the request gets held. Doing condor_release makes the jobs changing from the hold state into the idle state. The state does not change to run.
Why to the submitted jobs get held?
000 (011.000.000) 10/21 11:49:11 Job submitted from host: <10.110.44.83:49184>
012 (011.000.000) 10/21 11:49:28 Job was held.
Code 0 Subcode 0