Using Datahub in a Course
For students in a course that uses Datahub. The guide assumes only a web browser.
Start in the browserCoursework
Each guide is written for one role and states at the start what it assumes you know.
Course access comes from the roster. Students do not request anything. Instructors request the course workspace before the term starts.
For students in a course that uses Datahub. The guide assumes only a web browser.
Start in the browserFor students whose course needs SSH, launch.sh, batch jobs, VS Code, or a GPU class the course menu lacks. The guide assumes a terminal.
Start in a terminalFor instructors and TAs, from requesting a course to grading its notebooks. Most tasks need only a browser.
Set up a courseOutside a scheduled course
A project or research workspace is separate from any course workspace. It has its own eligibility, resources, and access dates. These guides assume you are comfortable in a shell.
For independent study, capstones, and club or team projects. The guide explains eligibility and the request form.
Request project accessFor graduate students, postdocs, undergraduates, and staff researchers working without a lab workspace.
Start your researchFor faculty arranging access for a group. IT Services staff carry out some of the steps.
Set up a labEligibility and access
Access comes from the course roster. When Access Starts & Ends says when enrollment changes reach PARASOL.
Request the course through the Specialized Instructional Computing Course Request. See Requesting a Course.
Submit the Independent Study Request with the project scope and what it needs to run. See Eligibility & Requesting Access.
Ask Research IT whether PARASOL fits the project, then follow Research on PARASOL or Setting Up a Research Lab. Support for research work is from rcd-support@ucsd.edu.
If you are not sure which applies, email parasol@ucsd.edu.
Before your first session
Containers run as you, without sudo. Install packages in your own environment. See Customizing an Environment.
The login node only launches jobs and moves files. Run your work in a session. See The Login Node.
Booked or on demand, a GPU session draws on your workspace's Service Unit budget. See Service Units & Budgets.
A session that sits idle long enough is stopped. Save work as you go. See What Ends a Session.