Page History

...

The information can be provided by the sinfo command. From psana nodes:

psslurm ~ # sinfo
PARTITION AVAIL TIMELIMIT NODES STATE NODELIST 
anagpu* up infinite 1 drain psanagpu107 
anagpu* up infinite 6 idle psanagpu[101-104,110,113]

...

There are 2 ways to submit a job on the cluster. The main way is by using the sbatch command to take full advantage of the computing power, and the other is to submit an interactive job. 

Sbatch

The commands specified in the script file will be ran on the first available compute node that fits the resources requested.

...

[omarq@psslurm conda 06:10:08]cat tst_script 
#!/bin/bash
#
#SBATCH --job-name=‘name’ # Job name for allocation
#SBATCH --output=‘filename’ # File to which STDOUT will be written, %j inserts jobid
#SBATCH --error=‘filename’ # File to which STDERR will be written, %j inserts jobid
#SBATCH --partition=anagpu # Partition/Queue to submit job
#SBATCH --gres=gpu:1080ti:1 # Number of GPUs
#SBATCH --nodes=1  # number of nodes.
#SBATCH --mail-user='username'@slac.stanford.edu # Receive e-mail from slurm
#SBATCH --mail-type=ALL # Type of e-mail from slurm; other options are: Error, Info.
#
srun -l hostname
srun python ExampleMultipleChaperones.py


[omarq@psslurm conda 06:10:11]sbatch tst_script 
Submitted batch job 200187

sbatch tst_scirpt

Srun

The srun command gets control of a node to run jobs interactively.  These can be useful for data exploration and significant software development.

The following are a few examples:

psslurm conda 07:29:39 srun -N2 -n4 hello.mpi 
Process 0 on psanagpu110 out of 1
Process 0 on psanagpu110 out of 1
Process 0 on psanagpu113 out of 1
Process 0 on psanagpu113 out of 1

Monitoring

To Prior to submitting a batch job it is recommended to check that jobs that exist on the system by using use the squeue command:

psslurm ~ # squeue 
             JOBID PARTITION     NAME     USER ST       TIME  NODES NODELIST(REASON) 
               187    anagpu ExampleM    omarq  R       0:04      1 psanagpu110 

...