User Responsibilities

Anunna is a shared system. Keeping it usable and fair for everyone depends on every user following a few basic responsibilities. This page summarises what is expected of you; the linked pages give the detail.

Use the system as intended

• Do not run heavy work on the login nodes — they are an access point only. Run computation as jobs through the scheduler; see Scheduler Overview (Slurm).
• Do not run I/O-intensive work — including writing log files — against your home directory. It is on a slower NFS filesystem; write job output and logs to Lustre instead.
• Request only the resources (CPUs, memory, time, GPUs) your job actually needs, so that capacity remains available for others. See the Fair Use Policy.
• Do not run jobs outside the scheduler. Logging in to a compute node is only for inspecting your own running jobs; see Compute Nodes.

Look after your data

• You are responsible for your own data. Keep your own copies of anything you cannot afford to lose — not every storage area is backed up. See Backup Policy.
• Clean up temporary data. Tidy /lustre/scratch and node-local /tmp when you are done, and move data you want to keep but are no longer using to archival storage; see Storage Systems Overview and Archival Storage.
• Handle confidential or personal data responsibly and in line with WUR policy.

Look after your account

• Your account is personal — do not share your credentials or let anyone else use your account.
• Report problems, suspected misuse, or security concerns to the HPC team through Support.

See also

• Policies and Terms of Use
• Fair Use Policy
• Login Nodes
• Backup Policy
• Support