I imagine a wide problem in using HPC systems is not knowing how much memory or walltime to allocate. I imagine this not knowing the right amounts (of memory especially) causes huge amounts of wasted resources on HPC systems. I know it does in my workflow.
SLURM returns MaxRSS/Elapsed value for each job after completion which batchtools could store somewhere at the completion of jobs.
I'm eventually imagining a nice function exported from future.batchtools or drake that could report on these somehow.
I imagine a wide problem in using HPC systems is not knowing how much memory or walltime to allocate. I imagine this not knowing the right amounts (of memory especially) causes huge amounts of wasted resources on HPC systems. I know it does in my workflow.
SLURM returns MaxRSS/Elapsed value for each job after completion which
batchtools
could store somewhere at the completion of jobs.I'm eventually imagining a nice function exported from future.batchtools or drake that could report on these somehow.