Nublado

Nublado implements a Jupyter notebook service in Kubernetes, a WebDAV file server that provides access to the same underlying POSIX file system, and a client for speaking to the Jupyter notebook service and the Labs it spawns, along with a set of mocks to be used when developing a service with that client. It is the software that provides the Nobebook Aspect service of the Vera C. Rubin Observatory Science Platform.

This site documents the Nublado service for platform administrators. Users of the Rubin Science Platform should instead see the documentation at rsp.lsst.io.

Nublado provides only the service to create and manage Kubernetes pods running JupyterLab, the WebDAV file servers, and a client for talking to the Jupyter service and the Labs it spawns (as well as a set of mocks for users of the client to use). Nublado’s Jupyter Service can spawn any compatible lab image, although it currently makes assumptions about the versioning scheme that are specific to Rubin Observatory. For the official Rubin images, see sciplat-lab.

• About
  • Architecture overview
  • Component diagrams
  • Sequence diagrams

• Administrators
  • Configure home directories
  • Running init containers
  • Set up Google Artifact Registry
  • Stop all running user labs
  • Delete a JupyterHub user session
  • Wipe the JupyterHub database
  • Configuration reference
  • Google Artifact Registry (GAR)

• REST API

• Client guide
  • Installing the Client
  • Using the Client
  • Sequence of Events
  • Interacting With The Lab
  • Mocks and Testing

• Developers
  • Architecture of Nublado controller
  • Development guide
  • Release procedure
  • Nublado development plan
  • Python internal APIs