ColdFront Implementation

Over the past several months, the OIT high-performance computing (HPC) teams have been working alongside the O’Donnell Data Science and Research Computing Institute to implement a platform for resource management and reporting on the two HPC clusters. The software, ColdFront, was selected to address several critical needs:

• Improve the efficiency of resource allocations to enhance overall performance for all users
• Expedite the creation of accounts and access to the clusters through automation and integration with various systems
• Enhance resource management for storage and compute resources, and
• Provide critical University reporting for HPC activities.

ColdFront is an application that helps manage resource allocations for projects running on the HPC clusters. It allows authorized individuals to create new projects and assign users to collaborate on those projects. Faculty can then request resource allocations, both compute and storage, required for their specific projects. ColdFront automates the provisioning of new users onto HPC, enabling them to start working and collaborating more quickly.

A new integration solution for academic classes allows instructors to request HPC resources for their courses. All students enrolled in the course will automatically be provisioned for access to the class HPC resources. Students will be required to accept an end user agreement before using the HPC resources. Any newly enrolled students will be added automatically, and any students who withdraw will no longer have access to the class HPC resources.

In addition to these new automations, new reporting tools are being installed and integrated with ColdFront to allow researchers access to data analytics regarding their projects and the usage of allocated resources. These metrics provide insights for faculty on their specific activities per project, as well as critical information to University leadership on the usage of this essential research instrument.