Importing a Docker or a Singularity prebuilt container requires resources that are limited on the login node, so first we need to get a terminal on a debug computing node. Find out the available projects:
module load sdsc # should be already loaded
expanse-client user -r expanse
Copy-paste the more suitable "Project" into an environment variable:
export PROJECT=xxx123
Now we can request 30 minutes in the debug partition:
srun --partition=debug --pty --account=$PROJECT --nodes=1 --ntasks-per-node=4 \
--mem=8G -t 00:30:00 --wait=0 --export=ALL /bin/bash
For example we can import one of the Docker containers by the Jupyter Project that can be used to run JupyterLab on Expanse via the Galyleo script:
cd /expanse/lustre/scratch/$USER/temp_project
module load singularitypro
singularity pull docker://jupyter/scipy-notebook:latest
This will create the Singularity container file:
$ du -sh scipy-notebook_latest.sif
875M scipy-notebook_latest.sif
We can get a terminal inside the container and verify it is an Ubuntu release:
$ singularity shell scipy-notebook_latest.sif
Singularity> cat /etc/lsb-release
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=20.04
DISTRIB_CODENAME=focal
DISTRIB_DESCRIPTION="Ubuntu 20.04.4 LTS"
Finally execute a command through the container with singularity exec:
$ python --version
Python 3.6.8
$ singularity exec scipy-notebook_latest.sif python --version
Python 3.10.4