SLURM

Official Docs

see partitions

sacctmgr show assoc user=$USER format=Cluster,Account,Partition,QOS

see nodes and thier configure cpus and gpus sinfo -o "%P %N %G"

job allocations

srun --account=YOURGROUP_NAME --nodes=1 --partition=es1 --time=1:0:0 --qos=FILLIN --pty bash

see your jobs

squeue -u <your_user_name> and scontrol show job <your_job_number>

see job statuses

sacct --format=jobid,jobname%40,state

Output logs will be save to current directory.

Monitoring

srun –jobid=$your_job_id –pty /bin/bash

scontrol show job <jobid> then ssh <nodeid>

Various places that use slurm and have docs

StonyBrook