JupyterHub is a powerful platform that provides a multi-user web environment for interactive computing using Jupyter Notebook and JupyterLab. It enables multiple users to have their own isolated and customizable Jupyter servers, all accessible through a standard web browser. This makes JupyterHub an ideal solution for:

• Collaborative Work: Teams can work on projects simultaneously in their own environments.
• Education: Instructors can provide individual computing resources for students.
• Research: Researchers can access and analyze data in dedicated environments.

Accessing JupyterHub

To access the NAS Systems Environment (NSE) JupyterHub platform, please submit a NAMS request. Once you have been granted access to JupyterHub, you can access it at: https://jupyterhub.nas.nasa.gov/

User Space

When you use the NAS Data Portal JupyterHub you will be given your own user space with about 49 GB of disk space. Note that this data gets deleted 24 hours after you end your session/shut down your environment.

Limitations

The Data Portal instance of JupyterHub has the following limitations due to the hardware setup:

• If you want to perform random reads on a file, we advise that you copy the dataset over to your local storage.
• All user space environments are wiped every 24 hours after shutdown. Please copy your data out of the environment after you are done working.

Getting Support

If your team requires more resources and you want to discuss possible funding upgrades, or if you require other assistance or encounter issues, please contact the NAS Control Room at support@nas.nasa.gov.

Available Resources

The NAS JupyterHub instance offers convenient access to several key resources:

• Data Portal Public Data Files: You can directly browse and utilize public datasets available through the data portal within your Jupyter environment.
• Pre-built Containers: A wide range of pre-configured software containers are available, including popular options for:
  • For machine learning and deep learning tasks.
  • PySci: For scientific computing with Python (likely including libraries such as NumPy, SciPy, and Matplotlib).

For additional information, see the following articles:

• Secure Setup for Using Jupyter Notebook on NAS Systems
• How to Set Up R Kernel in JupyterLab
• Machine Learning at NAS