This chart lets you install a single user instance of jupyterlab into a kubernetes cluster.
There are more elaborate solutions out there (see jupyterhub) , but sometimes this is all you need.
Additional python requirements can be installed with helm config (see pythonRequirements
).
You can inject additional python files (e.g. containing helper functions) using extraConfigMapFiles
For example:
extraConfigMapFiles:
my_utils.py: |
def print_hello():
print("hello")
Then use in any notebook like so:
from my_utils import print_hello
print_hello()
You can use s3 as a storage backend for your notebooks. See s3
in values.yaml
.
This chart was inspired by Deepak Sood's medium post Deploying Standalone JupyterLab on Kubernetes for Early Stage Startups.