The previously announced
transition of the Casper cluster’s scheduler from Slurm to the PBS Pro workload manager will begin with the addition of 64 new high-throughput computing (HTC) nodes during the March 9-10 scheduled maintenance downtime. The HTC nodes will be accessible only through PBS Pro. The other Casper nodes will be transitioned to PBS Pro over the next several weeks according to the following schedule.
Thursday, March 11 - New HTC nodes released in PBS; other nodes remain in Slurm.
Wednesday, March 24 - About half of the other Casper nodes will move to PBS.
- 7 CPU-only nodes
- 4 GP100 GPU nodes
- 2 4xV100 GPU nodes
- 3 8xV100 GPU nodes
Wednesday, April 7 - Remaining nodes transition to PBS.
CISL will provide documentation and training to help users transition their Casper scripts and workflows to PBS. Watch for more announcements coming soon.