August 13-15, 2019 CENTERWIDE Maintenance Downtime

Date of downtime: Tuesday, August 13 beginning at 7am -  through approximately 5pm Wednesday, August 14 (extended to Thursday, August 15)


Status Report 8/14 - 4pm:  Updates to the accounts systems are taking longer than anticipated.  CCR services will remain offline until this process completes.  We anticipate being back in production by 5pm Thursday, August 15.  


Approximate time of outage:  2 full days


Resources affected by downtime:

UB-HPC cluster (general-compute, debug, viz, largemem, and gpu  partitions) 

Industry cluster (compute, scavenger partitions)

Faculty cluster

Portals: WebMO, OnDemand



What will be done: 

  • Slurm scheduler upgrade
  • Operating system updates and reboot of all cluster nodes
  • All user accounts will now have a private group that is the same as their username.  See step 2 of this article


Jobs will NOT be held in the queue during the maintenance downtime.  Users will need to resubmit any jobs that did not complete before the downtime started.


If you have  any questions or concerns please e-mail ccr-help_at_buffalo.edu