About Cirrus

Cirrus is both a Compute Cloud and a Storage Cloud.

The Faculty of Science Research Cloud Cirrus 

  • designed as a tier II resource (clusters of machines designed to provide immediately available computing power for medium-scale storage and jobs)
  • uses virtualization to enable researchers to have full control over their research computing environment
  • allows them to benefit from sharing resources and using centrally managed hardware.

The aim is to merge many of the different local research clusters scattered throughout the Faculty into a single resource with reductions in cost and hardware maintenance being passed back to the researchers along with the benefit of being able to share unused resources.

Cirrus was initially funded by the Dean of Science with funding for maintenance, upgrades or increases in capacity provided by the researchers using the cloud.

The resources are administered by a combination of IST, Compute Canada and SRIT.

  • IST manages the physical hardware providing power, cooling and network access as well as taking care of warranty replacements, new machine installation etc.
  • Compute Canada manages the virtualization software which will enable researchers to startup and shutdown machines, clone existing machines, manage virtual networks and storage volumes etc. 
  • SRIT tracks how the researchers are charged for resources on the cloud and helps researchers create the virtual machines and access to data storage that will work for them.

Compute Canada also manage the tier I resources which many researchers use and some of these resources are also switching over to using virtualization. The research cloud will be managed using the same software and configured in a compatible way. This will enable future integration between the faculty research cloud and the national Compute Canada cloud so that researchers will not need to reconfigure, recompile etc. jobs to be able to take advantage of the large scale, tier I resources Compute Canada can offer. It is also hoped that this approach may work with other resource providers such as Cybera, Amazon and IST.

The cloud is currently comprised of:

  1. compute servers yielding 408 CPU cores with 5TB of RAM; approximately 15GB of RAM for each CPU core
  2. Performance Disk Storage
  3. Bulk Disk Storage – 1.07PB (1084 TB)
  4. GPU slots – 14 available for purchase