I'm using Jupyterhub and I've been in the habit of creating my own kernels
by dropping json files into this directory:
~/.local/share/jupyter/kernels
where the kernel looks like
~/.local/share/jupyter/kernels/env0
kernel.json
logo-32x32.png
logo-64x64.png
And the `kernel.json` looks like:
{
"argv": [
"~/.conda/envs/env0/bin/python",
"-m",
"ipykernel_launcher",
"-f",
"{connection_file}"
],
"language": "python",
"display_name": "env0"
}
This works just fine for just me. However, I'm working with my admin to
set this up for other users less inclined to mess with config files.
My desire is to establish a set of kernels that everyone will have access
to when they kick off their own notebook server via jupyterhub. But I also
want them to be able to add their own kernels if they wanted.
My first plan would be to create symbolic links in each users
`~/.local/share/jupyter/kernels/` directory to some globally accessible
directory with the appropriately specified kernels. They could then create
additional kernels on their own in the same directory.
However, this seems like there should be a configuration option that points
jupyter to a directory to search for kernels and that we could build upon
available kernels by looking in additional directories. So that in some
final state I could have a global config that points to where the globally
accessible kernels are and users can have additional ones in their own
directory.
Is my current plan sufficient?
Is my current plan good?
Are there configuration options that help me accomplish my goal?
Thanks,
Sean
--
You received this message because you are subscribed to the Google Groups
"Project Jupyter" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jupyter/553692d6-80f9-4c59-9f6a-93ee77e2e874%40googlegroups.com.
