.. _Using Ehlers:

Using Ehlers
============

.. _Ehlers login:

Login
-----

There is 1 login node:

============================================  ======================================================
Hostname                                      Node type
============================================  ======================================================
``ehlers-login.informatik.uni-stuttgart.de``  Ehlers login node
============================================  ======================================================

Host key fingerprint:

=========  ========================================================
Algorithm  Fingerprint (SHA256)
=========  ========================================================
RSA        ``SHA256:yGt6GD4f92SJSByfDPbCa4AOVBTYH3Sl0A01UczM7IY``
ECDSA      ``SHA256:ljmf74qRwQia7mp8mX27VkbqoA1XOJUJIG5XLmAWL1c``
ED25519    ``SHA256:TrGQzb576d2C41xESYwlY6iQOsXG88RSZDqk7BbfEig``
=========  ========================================================

.. _Ehlers submitting jobs:

Submitting jobs
---------------

When jobs are put on hold, check that you specified the partition where the jobs are supposed to run via the `-p` option.

Batch command:

.. code-block:: bash

    sbatch --job-name="test" -p cpu --nodes=1 --ntasks=4 --mem-per-cpu=2GB job.sh

Job script:

.. code-block:: bash

    #!/bin/bash
    #SBATCH --time=00:10:00
    #SBATCH --output %j.stdout
    #SBATCH --error  %j.stderr
    module load gcc/11.3.0 openmpi/4.1.4_gcc_11.3_cuda-11.7 boost/1.82.0_gcc-11.7_ompi-4.1.4 cuda/11.7
    srun --cpu-bind=cores ./pypresso testsuite/python/particle.py

Note also that the Ehlers cluster does not always cancel a job immediately if the resources that are requested are impossible to provide. Your job will instead sit in the queue with a small message saying impossible job.