SLURM Admin Guide — KFUPM JRCAI¶
Internal documentation for setting up and operating the KFUPM JRCAI HPC cluster: SLURM workload manager, LDAP authentication, NFS shared storage, user lifecycle automation, and day-to-day cluster operations.
Internal use only
This guide documents live infrastructure, including real hostnames, IPs, and configuration. It is for KFUPM JRCAI admins — do not share outside the team.
-
Installation
Build the cluster from scratch: NFS, LDAP, MUNGE, SLURM controller/worker, GPU support.
-
User Management
Account lifecycle, automated provisioning, SSH access control, disk quotas, and QoS budgets.
-
Operations
Day-to-day cluster administration: node state changes, conf redeploys,
clush. -
Troubleshooting
Known issues and how to fix them.
-
Hardware
Runbooks for physical infrastructure — storage, power, and other hardware procedures.