Introduction to the SCC

Boston University's Shared Computing Cluster (SCC) is a Linux cluster of more than 900 compute nodes, 25,000 CPU cores and 9 PB of storage. It is where BU researchers run the analyses that are too large, too slow, or too repetitive for a laptop.

This guide is for people who have never used a cluster before. It starts at logging in and ends with a real neuroimaging job running on the batch system. No prior Linux experience is assumed.

Before You Start

You need an SCC account, and every SCC account belongs to a project headed by a full-time BU faculty member. If you do not have one yet, start at the RCS account management page.

What This Guide Covers

Read the pages in order the first time through. Each one builds on the last.

• SCC OnDemand — the web portal, and what you can do from it.
• Connecting to a Desktop Session — getting a graphical Linux desktop on the cluster.
• The Desktop Environment — finding your way around once the desktop appears.
• The Linux Command Line — the terminal, the prompt, and the handful of commands you need.
• The Linux File System — home, /project, /projectnb, and file permissions.
• The Module System — loading software with module.
• The Batch System — submitting jobs with qsub.
• Customizing the Command Line — making the shell work the way you do.
• Yaxil Demo — pulling imaging data from XNAT, start to finish.

Additional Resources

• RCS research computing support pages
• Running jobs on the SCC
• Software packages on the SCC
• RCS training and tutorial materials

Questions? Contact us at help@scc.bu.edu.