[mvapich-discuss] PMGR_COLLECTIVE ERROR

[Andre Pattantyus]

Hello,

I am running a numerical dispersion model on a linux cluster with
mvapich-1.2.0 and  pgi/10.2 compiler. I am trying to submit a job on 8
processors with code written by the developer but I keep getting errors.
Right now I am getting the following error when the executable hycm_std
gets called:

/share/huina/akp4221/hysplit/trunk/exec/hycm_std
PMGR_COLLECTIVE ERROR: unitialized MPI task: Missing required environment
variable: MPIRUN_RANK


I think maybe my job submit command might not be adequate. Right now I am
submitting via this piece of code:

#PBS -q default
#PBS -N hysprun

# request X processes (X = nodes * cpus)
#PBS -l nodes=1:ppn=8

# request wall time
#PBS -l walltime=01:00:00

# PBS output files
#PBS -o $PBS_JOBNAME.OUT
#PBS -e $PBS_JOBNAME.ERR

echo "script running on host `hostname`"

# directory where qsub started
cd   $PBS_O_WORKDIR
echo "PBS_NODEFILE"
#cat  $PBS_NODEFILE
echo `pwd`

rm -f  MESSAGE.???

#-------------------------------------------------------
#  step2 - run the parallel or serial version of HYSPLIT

${EXECprod}/hycm_std        >>step2.sh

I do not have root privileges but I would like to know what the potential
issues I could be facing so I can let the administrator now. Mahalo!


-- 
Andre Pattantyus
Graduate Student Research Assistant
Department of Meteorology
University of Hawaii at Manoa
2525 Correa Rd, HIG 350
Honolulu, HI 96822
Phone: (845) 264-3582
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.cse.ohio-state.edu/pipermail/mvapich-discuss/attachments/20130626/4c725f4c/attachment.html