DCSC logo
 
ABOUT-DCSC
DCSC/SDU
DCSC/AU
DCSC/AAU
DCSC/DTU
DCSC/KU
 
+Open all         -Close all
 
 
#!/bin/csh
#
# PBS jobscript example for running a job on one node.
#
# A PBS batch job script is simply like a shell script - except
# that PBS will pick up commands disguised as comments in the first
# block of comments in the script, as shown here.
#
# The next line requests on node to be used for this batch job.
#
# Remember that the nodes in the cluster has multiple cores
# thus your application should be parallelized in some fashion and
# prepared to use all 4 cores per node.
#
#PBS -l nodes=1
#
# The next line requests that PBS sends an e-mail notification at the
# start of execution and at the end of execution.
#
#PBS -m be
#
# The next line, if one "#" is removed, will specify a maximum wall-clock
# limit of 1 hour, 10 minutes. If the time limit is exceeded the job will be
# terminated. The format is: HOURS:MINUTES:SECONDS.
#
##PBS -l walltime=1:10:00 
#
# The next line, if one "#" is removed, will specify that the job should be
# send to the 'workq' queue. 
#
##PBS -q workq
#

#
# Change to the directory where the qsub command was issued.
#
cd $PBS_O_WORKDIR

#
# Copy all files to be used in the run to /scratch
#
cp a.out input /scratch

#
# Change directory to /scratch - and run a.out. a.out is assumed to
# read the file 'input' and produce 'output'. You are REQUIRED to
# use /scratch for temporary files.
#
# We are of course assuming that a.out is suitable parallelized
# to run on multi-core systems.
#
cd /scratch
./a.out

#
# a.out has finished - move the output to $PBS_O_WORKDIR.
# Files can also be moved to /wkspace1 or /wkspace2.
#
mv output $PBS_O_WORKDIR

#
# PBS will erase ALL files in /scratch after the job is done.
#