Due to the shared nature of the FASRC cluster and accounts system, coursework and classes cannot be accommodated on the main FASRC cluster*.  We are unable to honor FERPA blocks and similar requirements on the FASRC cluster. Such information (names, usernames, etc.) is not visible outside the cluster, but it is potentially visible to other users on the shared cluster.

Academic Cluster

Classes in the FAS, SEAS and HCSPH can, however, use the FAS-OnDemand academic cluster, a cluster built in conjunction with the Harvard Academic Technologies Group which is very similar to the FASRC cluster but is dedicated to supporting academic coursework. Students can access this cluster directly from a link on the Canvas page of your course, and connect to an interactive computing portal (a dedicated instance of OpenOnDemand) which allows to run interactive computation, like Jupyter notebooks, Rstudio or Remote Desktop sessions, or to submit batch computation. Accounts are generated on the fly the first time a student connects from Canvas to FAS-OnDemand.

Any instructor who is interested in using FAS-OnDemand to run Jupyter notebooks, RStudio, Matlab or other applications on the cluster should refer to this form at ATG to request access for their class.

PLEASE NOTE: There is generally a 7-10 business day lead time for setting up new courses. This is especially true at the start of a semester when the majority of requests come in.

Students/Users

Please direct any questions or issues to Harvard Academic Technologies Group (atg@fas.harvard.edu)

* – If a PI/lab decides to sponsor regular accounts for users on their own for the purpose of a course, they should be aware that these users will have full access to their cluster lab group, storage, and fairshare. We do not recommend or support this. Please use the academic cluster or contact us if there is a compelling reason to do this.