Tutorial 5 (How to launch a quick calculation on waves nodes.)

More information here

The examples in this tutorial are only validated from the Nautilus Login node and not from guix-devel-001.

1. Where are the waves nodes?

from nautilus login node you can see the waves nodes with

sinfo --cluster=waves

The same QoS is applied to the Waves cluster.

glicid_qos

2. Work in the /scratch/waves space

If the calculation starts on Waves, you have to use /scratch/waves, which is optimal in these conditions. /scratch/nautilus is also available, but cannot benefit from the same access conditions as from cnode machines (non-native network and lower bandwidth).

In context of calculations, these penalties can cause /scratch/nautilus may not respond correctly due to the limitations, and this could impact jobs and even the nodes themselves.

regardless of the scratch program used, it is necessary to move the resulting data once the calculation campaign is complete.

As /scratch/waves is currently limited in space, please pay particular attention to its use. The available space will be gradually extended.

3. quickstart cpu slurm on waves.

#!/bin/bash

#SBATCH --job-name=myjob        # Name for your job
#SBATCH --comment="Run My Job"  # Comment for your job

#SBATCH --output=%x_%j.out      # Output file
#SBATCH --error=%x_%j.err       # Error file
#SBATCH --partition=standard      # partition standard
#SBATCH --time=0-00:05:00       # Time limit
#SBATCH --ntasks=1              # How many tasks per node
#SBATCH --cpus-per-task=1       # Number of CPUs per task
#SBATCH --mem-per-cpu=10g       # Memory per CPU
#SBATCH --qos=quick             # priority/quality of service

# Command to run
hostname                        # Run the command hostname

In this example, we run the bash command hostname.

3.1. To submit this job, run:

sbatch --cluster=waves myjob.slurm

This will submit your job to the Slurm for execution and a message with Job_ID will be displayed.

$ sbatch myjob.slurm
Submitted batch job 557978 on cluster waves

in myjob_557978.out you can see the hostname

if you want to scancel a job if your are on Nautilus login node you have to specify the waves cluster

Example:

scancel --cluster=waves 557978