GPC is Migrating to the SLURM Scheduler

Resource management and load balancing are controlled by GPC’s scheduler. Running a batch job on GPC begins with creating a wrapper script, followed by submitting to one of the queues via the scheduler. Previously GPC had used SGE as its job scheduler but as of January 2021, we will begin migrating to the SLURM scheduler. During this transition period both SGE and SLURM schedulers will be available, on non-overlapping nodes. The remaining SGE nodes will continue to be contained in the "all.q" and "himem.q" queues and the new SLURM converted nodes will be available in SLURM's "normal" and "highmem" partitions.

Please begin transitioning your jobs to SLURM for a smooth crossover when GPC becomes SLURM-only in the coming months. Most nodes will be converted to SLURM by mid-February 2021, with the final 2 nodes scheduled to be converted in May 2021. 

For SLURM job submission guidelines, please see our new SLURM documentation which includes a guide on converting SGE jobs to SLURM among other useful commands.

For legacy SGE instructions, please see our existing SGE documentation page.
 

SLURM Migration Schedule

Jan 11th:

• node01
• node05
• node06

March 22nd:

• node02
• node03

April 19th:

• node04
• node09

May 17th:

• node07
• node08