The changes to preemption should not have caused any difference in matchmaking between jobs and machines. To find out why jobs are not matching, you can get the list of requirements for all jobs by running:
condor_q -format "%s." ClusterId -format "%s: " ProcId -format " %s\n" Requirements
Unfortunately, condor_q -better-analyze (which gives more detailed information about why jobs are not matching) is not yet supported on Windows.
Also, restarting all machines does the equivalent of a condor_reconfig -all, so there is no reason to run that command after restarting.
Cycle Computing, LLC
Leader in Condor Grid Solutions
Enterprise Condor Support and Management Tools
On Feb 20, 2009, at 5:18 PM, Alas, Alex [FEDI] wrote: