jupyterlab-on-k8s

The jupyterlab-on-k8s project is a simple means of deploying jupyterlab on Kubernetes. In contrast to jupyterhub, jupyterlab is designed to support one user. Consequently, jupyterlab-on-k8s provides a highly straightforward means of testing Python client libraries against a variety of apps deployed within and outside Kubernetes--all within a familiar Jupyter notebook environment.

Python Version

As of 20211208, the Python versiuon for jupyterlab-on-k8s is 3.9.2. Consequently, all Python notebooks are v3.9.2 while all pip installs must be executed with the pip3 application.

Docker Build

export VERSION=0.0.1
export JUPYTER_VERSION=3.0.0

docker build --build-arg JUPYTER_VERSION=$JUPYTER_VERSION -f src/docker/jupyterlab-docker -t hokiegeek2/jupyterlab-k8s:$VERSION .

Helm Install

values.yaml Options

######################## Pod Settings ########################

releaseVersion: 0.0.1
imagePullPolicy: IfNotPresent

#################### Jupyterlab Settings #####################

app:
  namespace: default
  numServers: 1
  resources:
    limits:
      cpu: 500m
      memory: 512Mi
    requests:
      cpu: 250m
      memory: 256Mi

service:
  name: jupyterlab
  port: 8888
  targetPort: 8888
  type: LoadBalancer

Install Command

helm install jupyterlab deployment/charts/jupyterlab-on-k8s/