Run JupyterLab on Windows

This tutorial will guide you to run Jupyter Lab on MeluXina if your workstation is using Windows. It assumes you are using MobaXterm as an SSH/SFTP/X11 client.

  1. Connect to MeluXina with MobaXterm, for more details see the Windows/MobaXterm section in our connecting to MeluXina page.

  2. Allocate a cluster node on MeluXina with the salloc command

    • For example: salloc -A YOUR_ACCOUNT -t 0:05:0 -p cpu -q short -N 1
    • The example above will allocate one CPU node (-p cpu) in interactive mode, with the short QoS for a rapid allocation, for 5 minutes (-t 0:05:0).
  3. Activate Jupyter Lab

    • Load the JupyterLab module to be able to use it: module load JupyterLab
    • This will load the latest available version of JupyterLab. If you require a specific version, find if it is available in our general software stack using module available Jupyter, then load that specific version with `module load JupyterLab/VERSION'
  4. Start Jupyter Lab

    • Start Jupyter Lab: jupyter-lab --app_dir=/project/home/YOUR_PROJECT/ --no-browser --port=8888 --ip=0.0.0.0
    • This will launch a Jupyter Lab session which resides in your project workspace, with the port 8888 and able to accept connections from the MeluXina login node - this is needed for the next steps.

    start_jupyter_lab

  5. Create a local port forwarding configuration with MobaSSHTunnel

    • The notebook server is now started on a compute node and you will need to make a connection from it to your local computer through a private and secure tunnel, using a MeluXina login node as a gateway. MobaXterm offers an interactive interface where you can set up an ssh tunnel with ease:
    • (1) Open Tunneling
    • (2) Create a new connection
    • (3) Local port
    • (4) Address of login node
    • (5) Port of login node
    • (6) Username
    • (7) Cluster node (server name)
    • (8) Port of Jupyter Lab running on cluster node (server port)
    • (9) Import SSH private key

    open_tunneling open_tunneling_2 open_tunneling_3

  6. You are now ready to access the Jupyter Lab on your computer!

    • Open your workstation's browser and connect to

    jupyter_local