Work related to the forthcoming HPC system refresh is well underway. At this time we expect that to be able to refresh the system by the middle or late October.
What follows is a high level list of some of the things that will be accomplished during the upgrade period.
- Replace/upgrade the HPC's primary parallel filesystem with a new system with greater capacity and greater performance.
- Add/integrate new owner nodes (finished systems evaluations this week)
- Update Software Components (e.g., RHEL, MOAB/Torque, Compilers)
- Update Lustre file system (includes file quotas)
- Reconfigure scheduler to prioritize owner jobs (testing ongoing)
- Add new nodes to the Class/Test queue
- Add to and reconfigure core HPC network components
- Recompile core software components
More details will be sent to
to keep everyone up-to-date.