Paritions

The HPC SLURM queue has several available partitions. If no partition is specified the default is grit_nodes. The other partitions are to help manage groups with their own infrastructure that have joined the slurm queue. These partitions are named after the lab followed by _nodes, ie chc_nodes. If you are a member of a group with their own hardware you can enter multiple partitions and jobs will attempt to use the partitions in the provided order. This allows groups that are sharing their resources with the default queue to dip into the public resources if their private systems are busy.