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 behaviour that has not been verified against a primary source. Check with us before relying on anything here.

Datahub & DSMLP Documentation


This is the documentation set for UC San Diego's Datahub and Data Science & Machine Learning Platform (DSMLP). It is organized by audience and task, with the technical detail written once and linked from everywhere it is needed.

Please reach out to our team with any questions or feedback regarding these guidelines or Datahub/DSMLP as a whole.

Start Here


Each page states its assumptions in its opening lines.

Page Intended for Assumes
Overview Everyone; routes to the appropriate page Nothing
Using Datahub in a Course Students enrolled in a course that uses Datahub A web browser
Working from the Command Line The same students, when a course requires a terminal A terminal and ssh
Teaching with Datahub & DSMLP Instructors and TAs A browser; a shell for customization
Projects & Independent Study Student projects, independent study, capstones, clubs and teams Familiarity with a shell
Research on DSMLP Researchers working without a lab workspace Familiarity with a shell
Setting Up a Research Lab Faculty provisioning access for a group Some steps are performed by our staff

Technical Documentation


Detail that applies to more than one audience is documented once, in the subject area it belongs to, and linked from the audience pages.

Subject Covers
Access The browser, the login node, remote editors, sign-in problems, and when access starts and ends
Workspaces & Storage The unit that governs rosters, storage, images, GPU access and quotas — and where files live, how they move, and datasets
Running Jobs launch.sh, job modes and runtime limits, watching a job, checkpointing, and Kubernetes
GPU Access GPU classes, reservations, Service Units, quotas and availability, and what ends a session
Environments Standard images and pinning, customization, and building a custom image
Grading Grading tools and interfaces, the notebook workflow through to Canvas, and recovery
Reference Error messages, support routing, the glossary, HPC vocabulary, group management, policy and software