A Kubernetes Primer For Data Scientists, Part 1

I deploy JupyterHub on Kubernetes, and explain key concepts and show you how to be dangerous along the way. No prior Kubernetes experience is required.

Resources:

Why learning Kubernetes is a good idea for DS/ML people: https://hamel.dev/blog/posts/k8s/

Errata and additional notes:

18:18: I talk about `config.yaml` for the helm chart. config.yaml overrides the default parameters that are listed in `values.yaml` in the helm chart I showed at 15:38. In the video, I do not use the argument `–values config.yaml` because I do not want to override any of the default values. In practice, if you wanted to deploy JupyterHub in the enterprise, you would want to customize and configure things like security, hardware, and dependencies. I suggest looking at the customization guide: https://z2jh.jupyter.org/en/stable/jupyterhub/customization.html for an overview of the things you can customize in your JupyterHub deployment.