How To Check CPU Utilization In Slurm Cluster?

Use `squeue -u $USER` to see your running jobs and job IDs

Use `sstat -j ` to view CPU usage for a running job

Use `sacct -j –format=JobID,JobName,AllocCPUS,Elapsed,CPUTime,TotalCPU,MaxRSS` to inspect job CPU usage after completion

Use `seff ` to get job efficiency and CPU utilization summary

Use `scontrol show job ` to check allocated CPUs and job details

Use `top` or `htop` on the compute node to monitor per-process CPU usage

Use `ps -u $USER -o pid,ppid,cmd,%cpu,%mem –sort=-%cpu` to list your processes by CPU usage

Use `srun –pty bash` and then `top` or `htop` inside the allocated node for interactive monitoring

Use `sacct -j –format=JobID,AllocCPUS,CPUTimeRAW,TotalCPU,Elapsed` to compare allocated versus used CPU time

Use `scontrol show node ` to view node CPU availability and load-related details

Use `sar -u 1 5` if system accounting is enabled on the node

Use `mpstat -P ALL 1 5` if available to check per-core CPU utilization

Use `watch -n 1 “squeue -j “` to monitor job state changes while it runs