Skip to content

JupyterLab

A highly extensible, feature-rich notebook authoring application and editing environment

Homepage Version(s): 4.3.5

This application is available in Open OnDemand.

First, you must upload your data onto the HPC. Or, if your files are small, you can upload them directly using Open OnDemand

Accessing JupyterLab on Open OnDemand#

  1. Log into Open OnDemand
  2. In the top navigation bar, select "Interactive Apps" and then from the drop down, "JupyterLab".
  3. Select the Slurm Account/partition that you want to use for your job. If you leave this blank, the app will run in our genacc_q account/partition.
  4. Fill out the rest of the job submission form with your desired specifications.
    • Python Version: Specify the desired Python version for your job execution. This can be specified as either a Python version (e.g., python/2, python/3) or an Anaconda version on the HPC system.
    • Path to Conda Environment: Enter the full path to your existing Conda environment where Spyder is installed (e.g., ~/myApp). If such an environment does not exist at the given path, the system will generate one for your project. Your job will remain in Starting mode while the environment is configured. This can take up to 10 minutes.
    • For optimal performance during environment creation and subsequent Spyder job execution, it is recommended to allocate a minimum of 8 CPU cores and 16G of memory resources.
  5. Click "Launch" to queue your job.
  6. When your interactive job has started, click Launch.
  7. A new tab will open, and the Spyder Application will launch shortly after.

Note

More information on Launching Interactive Apps through Open OnDemand

Accessing Files Outside Your Home Directory#

By default, Jupyter sessions launched through Open OnDemand are restricted to the user's home directory. This limitation is due to Jupyter's internal configuration, which prevents access to directories outside the home path. Consequently, files stored in /gpfs/research or other locations outside your home directory cannot be directly accessed or browsed within Jupyter.

If you need to access your research storage from within Jupyter, you can create a symbolic link from your home directory to your research folder.

Workaround Steps#

  1. Open a terminal (either through Open OnDemand or SSH).
  2. Create a symbolic link from your research directory to your home directory using the following syntax:
1
$ ln -s /gpfs/research/<RESEARCH_VOLUME> /gpfs/home/<FSUID>/<LINKNAME>

For example:

1
ln -s /gpfs/research/scratch/abc25d /gpfs/home/abc25d/scratch

This creates a symbolic link called scratch in your home directory that points to /gpfs/research/scratch/abc25d.
Once created, you can navigate to scratch in Jupyter as if it were a normal folder, and access your research data.