The BIC is piloting JupyterHub: Multi-user server for Jupyter notebooks

The BIC is currently piloting JupyterHub to provide multi-user support for users of Jupyter Notebooks. Eventually this hub may have access to HPC resources available on the BIC’s compute cluster.

What’s a Jupyter Notebook?

The Jupyter Notebook is an open-source web application that allows you to create and share documents that contain live code, equations, visualizations and narrative text. Uses include: data cleaning and transformation, numerical simulation, statistical modeling, data visualization, machine learning, and much more.

There is a large collection of example notebooks provided by the Jupyter community at this link. It’s a good place to see the range of capabilities available with Jupyter.

What’s JupyterHub

JupyterHub brings the power of notebooks to groups of users. It gives users access to computational environments and resources without burdening the users with installation and maintenance tasks. JupyterHub runs in the cloud or on your own hardware, and makes it possible to serve a pre-configured data science environment to any user in the world.

Getting Access

  1. Access the JupyterHub at https://jupyter.bic.mni.mcgill.ca
  2. Login using your McGill email and password.

Note: Currently (9/21/18) there is an error with the Login that is not accepting some valid McGill accounts. This message will be updated once it’s fixed.

Note: during the JupyterHub pilot phase, the server may go down for extended periods. Please use this thread to report any issues.

Your help is needed during the pilot

While trying out Jupyterhub, please use this forum thread to provide feedback.

  • If you are unable to login with your McGill account, please report this. JupyterHub authenticates through McGill’s LDAP authentication system. Currently, a McGill account is required.
  • If you encounter any bugs or error messages, please take a screenshot and explain what led to the bug ocuring.
  • If there are any special features of Jupyter that you would like to see added, please tell us. For example, Jupyter now supports kernels configured for other analytic languages like R and Matlab.
  • If everything is working fine, please reply on the forum to tell us this. Otherwise we won’t know.

Thank you