Oct 2026 HPCF Newsletter

Happy end of October everyone!

I'm writing to give everyone an update regarding the UMBC HPCF and other Research Computing projects currently undertaken by DoIT during the month of October.

Summary

• Run Periodic Jobs on chip with Scrontab
• Slurm Backfill Scheduler Update
• Shared Infrastructure Governance Committees Reminder

Run Periodic Jobs on chip with Scrontab

During our recent maintenance window, we enabled Slurm’s scrontab feature. You can now set up automated, periodic jobs using a syntax that combines traditional Unix crontab scheduling with Slurm SCRON parameters (similar to SBATCH script parameters).

Please refer to our documentation for more information and usage examples: https://umbc.atlassian.net/wiki/spaces/faq/pages/2712731655/Scheduling+Recurring+Jobs+with+Scrontab 

Slurm Backfill Scheduler Update

Over the last couple of months, we’ve investigated reports of users’ jobs remaining queued despite idle nodes being available. We found that our Slurm backfill configuration was resetting at the start of each cycle without evaluating more than the first 500 jobs, effectively waiting to consider any jobs beyond the first 500 pending jobs. To address this issue and ensure fair access for users, we’ve updated the Slurm backfill scheduler to continuously process the full queue and capped the number of jobs evaluated per user. Users can now expect smaller or shorter jobs to get backfilled and scheduled sooner as idle nodes become available.

If you experience unexpected issues with chip, please submit an RT Ticket so we can investigate!

Shared Infrastructure Governance Committees Reminder

The next meetings of the SIG Subcommittees are posted as events to the HPCF myUMBC Group and shared below. See the full schedule at this link.

• SIG-CPU: Nov 13th, 2026 @ 1500ET in PHYS 221 or Webex

• SIG-GPU: Not meeting in November

• SIG-Cloud: Nov 11th, 2026 @ 1200ET. Webex only

• SIG-User Experience: Nov 9th, 2026 @ 1300ET in ENGR 125a or Webex

Publications
If you have any publications, presentations, theses, or other works that made use of the campus cluster, please submit an RT Ticket with bibliographic information so that we can accurately reflect this work in our records and on the HPCF Website. 

Need Help?
As always, please communicate any issues/questions to the Research Computing RT Queue (hpcf.umbc.edu > User Support > Request Help).

Get spooked.
Gregory Ballantine,
HPC System Administrator