Contact support if you have any questions or concerns not addressed here.
HSUper’s parallel file system BeeGFS is available on the login nodes only. It cannot be used for jobs!
HSUper’s CephFS mount is available on all nodes. All data for jobs must be stored on the CEPH file system (e.g. /ceph/scratch/cerebras). See csctl get volume for more information.
On the login nodes, a local SSD provides a shared local scratch storage. Your /home directory lies on the Cerebras login nodes also within this local disk and cannot be used for jobs. As space is limited, remember to delete files not needed anymore.