Difference between revisions of "Running With Slurm"
Jump to navigation
Jump to search
Line 36: | Line 36: | ||
brian@brems:~$ squeue | brian@brems:~$ squeue | ||
JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) | JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) | ||
− | 4 brems runmcnpx | + | 4 brems runmcnpx brian PD 0:00 1 (Resources) |
2 brems runmcnpx brian R 33:02 1 brems | 2 brems runmcnpx brian R 33:02 1 brems |
Revision as of 01:42, 1 April 2010
[Slurm https://computing.llnl.gov/linux/slurm/] is the queuing system used on Brems. It allows multiple users to put jobs into a queue and the system to negotiate running them optimally. To use slurm with MCNPX do the following: (beta instructions)
Create a script file like this one, substituting your MCNPX parameters:
brian@brems:~/work/maxwell/run15$ cat runmcnpx #!/bin/bash #number of processes to run: #SBATCH -n 8 DATAPATH=/opt/mcnpx/data/ srun /opt/mcnpx/v27b_64_mpi_i8_slurm/bin/mcnpx i=14MeV.i
Add your run to the queue:
brian@brems:~/work/maxwell/run15$ sbatch ./runmcnpx sbatch: Submitted batch job 3
Verify that your run is in the queue:
brian@brems:~/work/maxwell/run15$ squeue JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 3 brems runmcnpx brian PD 0:00 1 (Resources) 2 brems runmcnpx brian R 11:54 1 brems
You can check the output of your run:
brian@brems:~/work/maxwell/run15$ cat slurm-3.out mcnpx ver=27b ld=Tue Aug 18 08:00:00 MST 2009 03/31/10 19:03:50
************************************************************* * * * MCNPX * * * * Copyright 2007. Los Alamos National Security, LLC. * * All rights reserved. * * * * This material was produced under U.S. Government contract *
To cancel a queued or running job use scancel:
brian@brems:~$ scancel 3 brian@brems:~$ squeue JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 4 brems runmcnpx brian PD 0:00 1 (Resources) 2 brems runmcnpx brian R 33:02 1 brems