I came here looking for this! The last time I tried it in early 2017-12 it
was still "broken" with SLURM 17.11.0. Glad to see that it was fixed with
17.11.1 (and to know why). I've now got PAM limits being applied correctly
on my cluster. Thanks for the link, Andy.
Cheers,
On Fri, Dec 8, 2017 at
Hello,
I am operating a small cluster of 8 nodes that have 20 cores (2 10-core cpus)
and 2 GPUs each (Nvidia K80). To date, I have been successfully running CUDA
code where I typically submit single-cpu single-gpu jobs to nodes via slurm
with the cons_res and CR_CPU options.
More recently, I