I’m trying to figure out the reasoning for the .py files in the zero to jupyter helm charts, specifically for the database connectivity. Couldn’t this have been accomplished by injecting the variables into the deployment yaml?
I don’t understand your question yet, so I’ll just mention a few things that may be relevant.
We are mounting the .py files to the hub pod in order to not rebuild the image whenever we change those. That helps users who customize the hub image to have various additional packages etc installed not need to rebuild their image as often as well.
The jupyterhub_config.py is reading configuration from a k8s Secret resources that has been rendered by helm to include a lot of the relevant Helm chart configuration passed by the user, such as
Thank you for posting this question in the forum rather than creating an issue on github, and welcome to the Jupyter forum!!