SLURM - Simple Linux Utility for Resource Management

Functions:

• Provide tools for starting, executing, and monitoring jobs.
• Provide tools for job resource and access time requests.
• Mangage job queues.

Links:

• Research Computing Support Services Documentation
   
• Slurm version 16.05 Documentation
   
• sbatch Documentation
   
• The National Energy Research Scientific Computing Center
   
• A SLURM Quick Start Tutorial
   

[siegelj@tc-login ~]$ scontrol version
slurm 16.05.0
[siegelj@tc-login ~]$ sinfo -all
Wed Jul 27 08:18:49 2016
PARTITION AVAIL  TIMELIMIT   JOB_SIZE ROOT OVERSUBS     GROUPS  NODES       STATE NODELIST
TC*          up   infinite 1-infinite   no       NO        all     10        idle tc-node[156-164,166]
[siegelj@tc-login ~]$
------------------------------------------------------------------------------------------------------
[siegelj@tc-login ~]$ sinfo -all
Mon Sep 12 08:56:16 2016
PARTITION AVAIL  TIMELIMIT   JOB_SIZE ROOT OVERSUBS     GROUPS  NODES       STATE NODELIST
TC*          up   infinite 1-infinite   no       NO        all      1    drained* tc-node156
TC*          up   infinite 1-infinite   no       NO        all      9        idle tc-node[157-164,166]
[siegelj@tc-login ~]$
------------------------------------------------------------------------------------------------------
[siegelj@tc-login ~]$ sinfo -all
Fri Oct 21 07:45:15 2016
PARTITION AVAIL  TIMELIMIT   JOB_SIZE ROOT OVERSUBS     GROUPS  NODES       STATE NODELIST
TC*          up   infinite 1-infinite   no       NO        all      1          down* tc-node156
TC*          up   infinite 1-infinite   no       NO        all      4       drained tc-node[157,159-160,166]
TC*          up   infinite 1-infinite   no       NO        all      1         mixed tc-node161
TC*          up   infinite 1-infinite   no       NO        all      4        idle tc-node[158,162-164]
[siegelj@tc-login ~]$ squeue -u siegelj
             JOBID PARTITION     NAME     USER ST       TIME  NODES NODELIST(REASON)
              1501        TC   factor  siegelj PD       0:00      6 (Resources)
              1502        TC   factor  siegelj PD       0:00      6 (Resources)
              1506        TC   factor  siegelj PD       0:00      6 (Resources)
[siegelj@tc-login ~]$ scancel -u siegelj
[siegelj@tc-login ~]$ squeue -u siegelj
             JOBID PARTITION     NAME     USER ST       TIME  NODES NODELIST(REASON)
[siegelj@tc-login ~]$
[siegelj@tc-login ~]$ date
Thu Nov 10 14:25:29 CST 2016
[siegelj@tc-login ~]$ sinfo all
PARTITION AVAIL  TIMELIMIT  NODES  STATE NODELIST
TC*          up   infinite      2    mix tc-node[159-160]
TC*          up   infinite      2  alloc tc-node[157-158]
TC*          up   infinite      5   idle tc-node[161-164,166]
[siegelj@tc-login ~]$