-
Notifications
You must be signed in to change notification settings - Fork 1
Using the HPC Scheduler
Anthony Dick edited this page Jun 12, 2019
·
3 revisions
The HPC has migrated from LSF to Slurm for scheduling jobs.
To use Slurm,
Load the module
module load slurm
Run the job
- From the command line:
srun -J myjobname -c #CORE -e ERROR_FILE -o OUTPUT_FILE -q PQ_dcnlab -p investor my_command_line_arguments
- Using a script:
#!/bin/bash
#---Number of core
#SBATCH -c 1
#---Job's name in LSF system
#SBATCH -J example_job
#---Error file
#SBATCH -e example_job_err
#---Output file
#SBATCH -o example_job_out
#---Queue name
#SBATCH -q PQ_nbc
#---Partition
#SBATCH -p investor
##########################################################
# Setup envrionmental variable.
##########################################################
export NPROCS=`echo $LSB_HOSTS | wc -w`
export OMP_NUM_THREADS=$NPROCS
. $MODULESHOME/../global/profile.modules
module load fsl
module load afni
##########################################################
##########################################################
echo "hello world!">