GT Home : : Campus Maps : : GT Directory

PACE clusters ready for research

This entry was posted by on Saturday, 10 February, 2018 at

Our February 2018 maintenance ( is complete ahead of schedule.  We have brought compute nodes online and released previously submitted jobs. Login nodes are accessible and data available. As usual, there are some straggling nodes we will address over the coming days.

Our next maintenance period is scheduled for Thursday, May 10 through Saturday, May 12, 2018.

– Both pace1 and pace2 GPFS systems now apply a limit of 2 Million files/directories per user. Please contact us if you have problems with creating new files or updating existing ones, or see messages saying that your quota is exceeded.
– We performed several maintenance tasks for both pace1 and pace2 systems to improve reliability and performance. This included rebalancing data on the drives as recommended by the vendor.
– Temporary links pointing to storage migrated in the previous maintenance window (November 2017) are now removed. All direct references to old paths will fail. We strongly recommend Math and ECE users (whose repositories are relocated as a part of storage migration) to run tests. Please let us know if you see ‘file not found’ type errors referencing the old paths staring with “/nv/…”
– Deletion of old copies of bio-konstantinidis and bio-soojinyi are currently pending, but we will start deletions sometime after the maintenance day.
– CNS users are are migrated to their new home and project directories.
– We completed all power work as planned.
Rack/Node maintenance
– To rebalance power utilization, a few ASDL nodes are moved and renamed. Users of this cluster should not notice any differences other than hostnames.
– VM servers received a memory bump, allowing for more capacity
– Recabling and reconfiguration of IB network is complete
– All planned Ethernet network improvements are complete
As always, please contact us  ( if you have notice any problems.


Comments are closed.