JupyterHub Quick Start Guide

DATA311 | Nicholas James Uhlhorn, Fall 2025

Western hosts instances of JupyterHub on both the CSCI cluster and the CSE cluster. The official docs can be found here. The purpose of this document is as a quick start guide.

Prerequisites

If you are not connected to Western’s WiFi you will need to be connected to the VPN, instructions for all major OS platforms can be followed here.
To access the CSCI cluster (please use this over the CSE cluster) you will need to have a CS account, the credentials are the same as your universal account and you should already have one from being enrolled in this course.

Connecting to JupyterHub

Access JupyterHub using this link: https://csci-head.cluster.cs.wwu.edu/ You should be met with the following screen, login using your universal account. Note: log in using your WWU username without the @wwu.edu; it may or may not work with it, but it should definitley work without.

WWU JupyterHub Login Screen showing WWU logo and input fields for both a username and password.

You will then be promoted with a server creation prompt if you do not currently have one running. Servers on the cluster run for up to 24 hours after your launch them, and time out after 1 hour of inactivity.

In the “Environment” drop-down menu, select DATA311. If you forget to do this, some necessary packages may not be installed and you’ll need to restart your server using the correct environment.

Unless you will need more resources, it is recommended that you either keep the defaults, or set:

Number of CPUs to 2
Amount of Memory to 4
and keep the other values at their defaults for now.

JupyterHub Server Options Screen with the following fields: “Number of CPUs (Max: 8)”, “Amount of Memory (in GB - Max: 32GB)”, “Number of GPUs (Max: 4)”, “Environment”, “Extra HTCondor options”

After you have set the options click the start button. You will see the following setup screen. Be patient as the server starts and you will eventually be redirected to the JupyterHub environment.

JupyterHub Server startup screen, consists of a loading bar and expandable event tab
JupyterHub workspace screen, includes a file explorer and empty workspace promoting user to create a new file or start a terminal, which includes python options that are relevant to this course.
You can now upload files to your cluster workspace by either clicking the upload button, located to the right of the blue plus (new launcher) button, or by dragging and dropping the file into the file explorer.

Shutting down your Server

When finished working, please remember to shut down your server; it will time out after an hour, but it’s better to free the cluster resources for other users when you’re done.

In Jupyter’s File Menu, select Hub Control Panel.
In the window that opens, click the red Stop My Server button.

Notes

Even though the server only lives for 24 hours and times out after 1 hour of inactivity, your notebooks won’t be deleted when the server is shut down. However, you will need to rerun the code cells whenever you need to start a new server.
By default the server will auto save your work every 2 minutes. If your connection is not slow, you may want to decrease this value to make sure your work is saved in case you get disconnected from the server unexpectedly.