Reaching Others University at Buffalo - The State University of New York
Skip to Content

Important Notices

Substantial system upgrades, installation of new hardware and other changes will take place at CCR over the next few months and throughout 2015.  The first, and most noteworthy, being the installation of a new 3 Petabyte high performance file system that should have more than 20 times the performance of our existing file system (in addition to 5 times greater storage capacity).   

Installation of the new storage system along with the other changes that are summarized below will necessitate downtimes at CCR in which our production systems (clusters, networking, and storage) will be offline.   We are doing everything we can to schedule these with a minimum disruption in our services.
The regularly scheduled monthly downtime for the UB-HPC cluster has been delayed.  Large scale testing of our new storage system will be conducted after the installation is complete.  The schedule was slightly delayed due to delivery issues during the recent winter storm.  The new downtime is tentatively scheduled for Friday, December 19th.  On Wednesday, December 17th some racks of nodes will be offline for the upgrade of network switches.  Nodes will be brought back into the pool as soon as the switches are replaced.  Thank you for your patience.
All CCR resources will be unavailable for use Tuesday and Wednesday, January 6-7, 2015 as we migrate to the new GPFS storage solution.  This includes all servers that mount the CCR storage area network (SAN):
The general compute and debug partitions of the CCR cluster (Rush) will be unavailable for use Tuesday, October 7th from 7am until approximately 12pm.   Jobs will be held in the queue until maintenance is complete.  If you have any questions or concerns please e-mail

Wednesday, March 7, 2012 ~ New "scrubber" script to begin running on /panasas/scratch.  All files over one month old will be deleted as per our scratch storage retention policy.