Historically check jobs with:
sacct -u $USER -S YYYY-MM-DD -o jobid,submit,start,end,elapsed,state
Note: instead of a date you can supply now-24hours to -S
To get a list of all format fields type:
sacct --helpformat
For current queued and running jobs, you can find the job with:
squeue -u $USER -l
To obtain information about a running job, such as the running node, type:
jobinfo -v JOBID
To examine the resource usage of the job, ssh into that node with:
ssh COMPUTE_NODE
Then run htop. See https://htop.dev/
For GPU jobs, you can use nvtop and nvidia-smi