When I check the SLURM queue or my job status, I see:

"Request Node not available, Reserved for Maintenance"


sinfo shows:

ReqNodeNotAvail, Reserved for maintenance



Monthly maintenance downtimes are scheduled the last Tuesday of every month (unless otherwise noted) on the Academic (ub-hpc), Industry (presto), and Faculty clusters.   Any jobs in the queue that can't finish before 7am the morning of the downtime won't run.  They will remain in the queue until the nodes are released back into production (usually by 5pm).


If you know your job could finish before the downtime begins, you can change the requested walltime in your script.  The default on both clusters is 24 hours and this is used if not specified in your SLURM script.  To specify a walltime other than 24 hours, add the following to your SLURM script:


#SBATCH --time=00:15:00

NOTE: time is in the format of  HOURS:MINUTES:SECONDS