[Condor-users] MPI Jobs strange problem

["Chaitanya V. Hazarey" <chaitanya.hazarey@xxxxxxxxx>]

Hi all,

This MPI problems do not seem to end. See I had got all the things 
working previously. But now out of the blue, I get this error in the 
SCHEDD log file that


9/4 00:18:13 Cluster 1862 is in idle_clusters, but no longer in job 
queue, removing
9/4 00:18:13 Cluster 1862 is in idle_clusters, but no longer in job 
queue, removing
9/4 00:18:13 Found no idle dedicated job(s)
9/4 00:18:13 No idle dedicated jobs, handleDedicatedJobs() returning
9/4 00:18:13 Entering DedicatedScheduler::checkSanity()


and



9/4 00:17:41 Started timer (19363) to call handleDedicatedJobs() in 2 secs
9/4 00:17:41 JobsRunning = 0
9/4 00:17:41 JobsIdle = 0
9/4 00:17:41 JobsHeld = 0
9/4 00:17:41 JobsRemoved = 0
9/4 00:17:41 SchedUniverseJobsRunning = 0
9/4 00:17:41 SchedUniverseJobsIdle = 0
9/4 00:17:41 N_Owners = 1
9/4 00:17:41 MaxJobsRunning = 200





and when I submit 100 jobs 1 seems to run and the rest sit idle there.


Please if any one could atleast guess the cause. It would be great. I 
have put all the machines in the TESTING_MODE setting so that they 
always run the jobs. The strange thing is that when I googled for this 
error I could not find a single matching query.




Thanks,


Chaitanya V. Hazarey


School of Technology and Computer Science,
Tata Institute of Fundamental Research,
Colaba, Homi Bhabha Road,
Mumbai, Maharashtra, 400005