Skip to content

Latest commit

 

History

History
25 lines (19 loc) · 1.12 KB

README.rst

File metadata and controls

25 lines (19 loc) · 1.12 KB

jupyterhub-docker

Note that the Confluence Page on the JH server has more detailed notes on working with the hub.

What is JH?

This repository is intended for use on an Ubuntu EC2 instance running Docker in the 4DN AWS Cloud (4dn-dcic). It will not function outside of this account as the resources it accesses are specific to that account. The hub allows authenticated and lab associated users to access Python notebooks that give more convenient access to the 4DN metadata. They are able to interact with both the metadata and the raw files themselves on S3 via a custom mounting strategy. User notebooks are also backed up on S3, and some example notebooks are provided.

Structure

See JH documentation on how JH itself works internally. We build a custom hub from Dockerfile.jupyterhub and singleuser container 4dn-docker-v2/Dockerfile. The hub launches our custom singleuser notebook containers when users log into the hub. Their user access keys are automatically sourced and used. Auth0 authentication through the portal is also used for login.