walltime 200
| Author | Message |
|---|---|
|
Written on: 24. 10. 2012 [12:19]
|
|
|
hhoffmann
Holger Hoffmann
Topic creator
registered since: 21.12.2011
Posts: 30
|
Hi, habe gerade folgendes bekommen: "Walltime value too high for the all queue! The maximum value is 200 hours (-l walltime=200:00:00) qsub: Your job has been administratively rejected by the queueing system. qsub: There may be a more detailed explanation prior to this notice." Das Jobfile hierzu sah so aus: ----------------------------------------------------------- #!/bin/bash -login #PBS -N M3_dayL_sp5_UBA #PBS -M hoffmann bgt.uni-hannover[dot]de#PBS -j oe #PBS -l nodes=1:ppn=1 #PBS -l walltime=00:10:00 #PBS -l mem=1gb #PBS -t 1-1736 # show which computer the job ran on echo "Job ran on:" $(hostname) # change to work dir: cd $PBS_O_WORKDIR model='M3_dayL' Run='UBA' variety='sp' pp='5' #PBS -v variety,pp,Run,model # log file name LOGFILE=$(echo $PBS_JOBID | cut -d"." -f1).log #Setup MCR cache directory locally export MCR_CACHE_ROOT=/tmp/mcr_cache_$PBS_JOBID mkdir -p $MCR_CACHE_ROOT # the program to run /home/nhpghoff/prj/fpr4/distrib/run_fpr4.sh /sw/aws/numerics/matlab/MATLAB_R2012a &> $LOGFILE #Remove temporary MCR cache directory /bin/rm -rf $MCR_CACHE_ROOT ----------------------------------------------------------- Natürlich sind 1736 jobs a 10 Minuten mehr als 200 h, aber bis vor 5 min war das kein Problem, da stets immer max 72 Stück gleichzeitig gelaufen sind. Wieso geht dies nicht mehr bzw. bleibt das so? Bin gerade voll drauf angewiesen ... Gruß, Holger [This article was edited 1 times, at last 24.10.2012 at 14:11.] |
|
Written on: 24. 10. 2012 [15:06]
|
|
|
hhoffmann
Holger Hoffmann
Topic creator
registered since: 21.12.2011
Posts: 30
|
Hi, 15:05 - submission does work again. Thx, Holger |
|
Written on: 24. 10. 2012 [15:31]
|
|
|
gerdes
Andreas Gerdes
registered since: 14.09.2010
Posts: 50
|
Hi Holger, indeed We returned to the earlier working version of our submit filter. The issue you stumbled upon is no longer present. We will fix this in our testing branch. Andreas |


bgt.uni-hannover[dot]de