Knowledge Base

JupyterHub on RACC2

Introduction

JupyterHub is a shared, web-based platform that gives each user access to their own private Jupyter environment on a central server. It provides access to JupyterLab, a modern interface for working with code, data, and documents — including Jupyter Notebooks, which are interactive documents used for data analysis, scientific computing, and software development.

JupyterHub on the RACC2 removes the need to manually start a Jupyter Notebook or JupyterLab session and then open it in a web browser. Instead, you can use a browser on your local machine (currently only supported via NX) to log in to a persistent web portal.

Under the hood, sessions are transparently scheduled on the interactive login nodes, in the same way as SSH login sessions, with the same generous resource limits: 12 CPUs and 256 GB of RAM per user session. Please be considerate with memory usage. The memory limit is intended to prevent an individual user from accidentally overwhelming the server, not to ensure strict fairness or protect against total memory exhaustion.

 

JupyterHub on RACC2 via NX

Use the following procedure to start a JupyterHub session on the RACC2:

  1. Start an NX managed desktop session.
  2. Open a web browser such as Firefox. You can find these under Applications > Internet in the top left menu.
  3. Navigate to the URL: https://racc.rdg.ac.uk/jupyterhub

Kernels:

In a JupyterLab session launched via JupyterHub, you will find kernels corresponding to relevant centrally managed Conda environments. A kernel is the backend process that runs your code in a notebook. Each kernel is tied to a programming language and software environment. You can also register your personal environments as kernels.

Instructions for creating personal Conda environments can be found here: https://research.reading.ac.uk/act/knowledgebase/python-on-the-academic-computing-cluster/

 

To register your personal Conda environment in JupyterHub:

    1. Open a Terminal from the JupyterHub interface:
      File > New > Terminal
    2. Activate your environment, for example: 
      module load anaconda
source activate myenv
    3. Install the Jupyter kernel package in your environment (if not already installed): 
      conda install ipykernel
    4. Register your environment as a Jupyter kernel: 
      python -m ipykernel install --user --name myenv --display-name "myenv"

    After completing these steps, your personal environment will appear in the Kernel > Change Kernel menu in JupyterLab under the name “myenv”, and as an icon to start a Notebook, or Jupyter Console (you might need to reload your session).

R

There is also an R environment available within JupyterHub. This is the same R environment which is available on the RACC2. R packages can be installed in your home directory. Advanced users might also use Conda to create multiple R environments, and install dependencies needed to build R packages. However, we are not routinely supporting R on JupyterHub at present.

 

 

 

DTS Status Page
Drop-in Sessions

ACT services access flow

News
Suggest Content…
Click here and let us know!

Related articles

Transferring files to and from Linux storage

NAG library

ssh iconExternally visible ssh server – arc-ssh

NX authorization code

Recover a file on research storage

Using R and R Studio on RACC

ssh iconAccess to ACT Servers via an SSH Client

Running Matlab Scripts as Batch Jobs

Cronjobs on the Academic Computing Cluster

Python on the Academic Computing Cluster