Notebook server configuration using the Google Cloud SDK

This package provides Python classes that can be used in the Jupyter config file (e.g. ~/.jupyter/jupyter_lab_config.py) in order to fill in some configuration using the Google Cloud SDK's gcloud tool.

Included features

This package provides utility methods to look up any configuration options stored in the active gcloud config, in particular the project and region.

Additionally, this provides a utility method to update a given Config object to connect to a kernel gateway URL managed by Google.

Prerequisites

Install both Jupyter and gcloud.

For the kernel gateway feature, you will need an installation of Jupyter that uses the jupyter_server project and the version of jupyter_server you have installed needs to be at least version 2.4.0.

You will also need to log in to gcloud:

gcloud auth login

... and configure your project and region:

gcloud config set core/project ${PROJECT}
gcloud config set compute/region ${REGION}

Install

Clone this repository, and from this directory run the following:

pip install .

Setup

If you do not already have a Jupyter config file (e.g. ~/.jupyter/jupyter_lab_config.py), the first generate one with the following command:

jupyter lab --generate-config

The open your config file and add the following two lines to the end:

import google.cloud.jupyter_config
google.cloud.jupyter_config.configure_gateway_client(c)