...

Due to the nature of SAS and old library dependencies, SAS can only run in a containerized environment through single compute node job submissions and does not support multi-node submissions.

Using SAS on the cluster

...

 sbatch sasMP.sh

Multithreaded GUI job

To submit a job that uses 10 computational threads on one node, create a submission script sasMPGPU.sh:

#!/bin/bash
#SBATCH --nodes=1
#SBATCH --ntasks=10
#SBATCH --partition=general-gpu
#SBATCH --output=outputfile.txt
#SBATCH --error=outputfile.txt
#SBATCH --gres=gpu:1

module load apptainer
apptainer exec --nv --unsquash -H $HOME:/home -B /gpfs/sharedfs1/admin/hpc2.0/apps/sas/9.4:/gpfs/sharedfs1/admin/hpc2.0/apps/sas/9.4 SAS.sif /gpfs/sharedfs1/admin/hpc2.0/apps/sas/9.4/SASFoundation/9.4/sas myprog.sas

Then submit the script by:

 sbatch sasMPGPU.sh

GUI/Interactive SAS use with SLURM

...

To run an interactive SAS session with GUI functionality, you should "ssh -XY" to the cluster from a Linux machine or macOS , macOS machine with X11 enabled, or a Windows machine with X11 enabled. For more info, check out our guide to using X11 on the HPC here. Then, run the below commands.

To open an interactive Matlab SAS window with 10 cores available, you will want to do the following:

[netidhere@login6 ~]$ srun --x11 -N 1 -n 10 --pty bash
srun: job 4261315 queued and waiting for resources
srun: job 4261315 has been allocated resources
[netidhere@cn528 ~]$ module load apptainer
[netidhere@cn528 ~]$ apptainer exec --unsquash -H $HOME:/home -B /gpfs/sharedfs1/admin/hpc2.0/apps/sas/9.4:/gpfs/sharedfs1/admin/hpc2.0/apps/sas/9.4 SAS.sif /gpfs/sharedfs1/admin/hpc2.0/apps/sas/9.4/SASFoundation/9.4/sas

SAS should load all the needed SAS windows.

The following screenshot gives an example on how SAS should look:

...

When browsing for SAS files under a directory, feel free to use the File Drop down list and navigate to the needed directory where the SAS files are located.

...