Login Sign up

Monday Oct. 12, 2020, 4:30 p.m.–Oct. 12, 2020, 4:45 p.m. in Jupyter Core

Running non-jupyter applications on JupyterHub with jupyter-server-proxy

Yuvi Panda

Audience level:
Intermediate

Brief Summary

JupyterHub can provide not just Jupyter Notebooks through your browser, but also other complementary tools - like RStudio, Visual Studio Code, or even a full blown Linux Desktop! jupyter-server-proxy lets you run all these tools in the same JupyterHub, providing an experience similar to having them on your local computer. Come for the demos, stay to learn how you can run this on your JupyterHub.

Outline

You might use a combination of tools when teaching / doing data science. RStudio is very heavily used by the R community, while Visual Studio Code is now the IDE of choice for many folks. And sometimes, you just need a full blown Linux Desktop - maybe your work needs (an GUI application](http://ds9.si.edu/). Or you just want to run Firefox inside Chrome...

Thanks to a very open & extensible design, JupyterHub can host all these applications (and more!), using the same infrastructure it uses for running Jupyter Notebook / JupyterLab. jupyter-server-proxy is a Jupyter Notebook extension that allows any web based application to run behind JupyterHub, benefitting from its authentication and spawning mechanisms. All these applications will run in the same environment, share the same home directory, and generally give you the experience of having them all installed in your local computer, but via a web browser!

We will show a few demos describing what is currently possible, and go through the process of getting an arbitrary application of your choice working inside JupyterHub, to get a better understanding of how and why this works. Finally, we will talk about some of the limitations of this appraoch, and where tools other than JupyterHub are the way to go.