nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqf5lrtj7q92vwa4k0v43l4duttulyuvw8epxdf37ncw33jyfjey9qhnyery (nprofile…yery) basically i'll be using slurm to manage contention for resources like vGPUs and external hardware-in-loop instruments but not actually having it directly launching CI jobs.
the CI runner will sbatch a bunch of job requests and then the slurm script will spawn the VM that does the real work
