Science SLURM compute cluster: using idle nodes

This post is over 5 years old, it may be out of date.

The SLURM configuration of the cn-cluster has been changed. This gives jobs of owners high priority in their own partition (on their own nodes) and it makes it possible for all users to use all nodes with low priority. Next to that, all users can submit jobs to run on the C&CZ purchased node, with shorter running jobs getting higher priority.