Good morning, My last few messages were so successfully addressed, that I've come back for more. This is a great list.
In our setup, PIs get a monthly allocation and users associated with those accounts that use our normal partitions and QOSs can no longer run once they've consumed those units until the next month, when the limit is reset. I know how to set all this up in slurm and have done so. It works nicely. But I'd like to add another queue (partition) that is has a lower priority and is preemptable that users can use even if the account has hit its limit for the month. I'd thought the way to do this was using the TRESBillingWeight parameter for the partition, but this only seems to partially do what I want. If there is still some time available, a user can use that queue and run a job just fine (it doesn't appear to get killed when the account passes its limit); however, if the hours are maxed out, slurm still wont deploy the job because resources are expended. Moreover, I'd like to use "safe" account enforcement ... which also appears to check even when the billing weight is 0'd out. So I've a feeling I'm just going about this the wrong way. What is the "correct" way to do this in slurm? Thanks, Paul.
