Draft

This documentation is in draft and under active review. Figures for the Fall 2026 GPU reservation system are not yet published, and some pages describe behavior that has not been verified against a primary source. Confirm any figure with datahub@ucsd.edu before relying on it.

Environments

This section covers the software images a session runs, adding packages to an environment, and building a custom image.

Page Covers
Standard Images, Tags, and Pinning The standard images, image inheritance, image names and tags, image selection for a session, the quarterly rebuild, and pinning a workspace.
Customizing an Environment What can be installed, Python packages in a private kernel, R packages, root access and system packages, and recovering a broken environment.
Building & Publishing a Custom Image Choosing a base image, the Dockerfile, building and publishing, course images, and testing a custom image on DSMLP.

← Documentation index