feat: support for launching an MAPDL instance in an SLURM HPC cluster

[germa89]

Description

In this PR, add support to launch MAPDL instances in an HPC cluster. At the moment only SLURM scheduler is supported.
We only run Ansys in the job, so the cluster does not need to support Python at all.
Because of how PyMAPDL-MAPDL interaction has been designed, you can use PyMAPDL methods to upload/download files from the MAPDL instance. You could also execute OS commands using mapdl.sys, but this NOT recommended.

Usage

We allow the user to specify the argument launch_on_hpc when using launch_mapdl:

from ansys.mapdl.core import launch_mapdl

mapdl = launch_mapdl(launch_on_hpc=True)

Under the hood, we "wrap" (literally) the command used to launch MAPDL on a sbatch command:

sbatch --wrap='%PYMAPDL_MAPDL_LAUNCH_CMD%'

where %PYMAPDL_MAPDL_LAUNCH_CMD% is something like /ansys_inc/v242/ansys/bin/ansys242 -j myjob -np 4 -grpc.

Additionally we provide the option for specify scheduler arguments using the argument scheduler_args:

scheduler_args = "-N 10" # it could also be a dict: {"-N": 10}
mapdl = launch_mapdl(launch_on_hpc=True, scheduler_args=scheduler_args)

In the above case, it is interesting to set the number of CPUs to the real value of allocated CPUs:

# the double minus is added for long commands (len(cmd)> 1).
scheduler_args = {"nodes":  10, "ntasks-per-node": 2} 

mapdl = launch_mapdl(launch_on_hpc=True, scheduler_args=scheduler_args, nproc=20)

The values of each argument are wrapped in single quotes (') to avoid parsing issues.

Issue linked

Related to #3467

Checklist

• I have visited and read Develop PyMAPDL.
• I have tested my changes locally
• I have added necessary documentation or updated existing documentation.
• I have followed the PyMAPDL coding style guidelines.
• I have added appropriate unit tests that also ensure the minimal coverage criteria.
• I have reviewed my changes before submitting this pull request.
• I have linked the issue or issues that are solved to the PR if any.
• I have assigned this PR to myself.
• I have marked this PR as draft if it is not ready to be reviewed yet.
• I have made sure that the title of my PR follows Commit naming conventions (e.g. feat: adding new MAPDL command)

[wiz-inc-572fc38784]

Wiz Scan Summary

Scan Module						Total
IaC Misconfigurations	0	0	0	0	0	0
Vulnerabilities	0	0	0	0	0	0
Sensitive Data	0	0	0	0	1	1
Secrets	0	0	0	0	0	0
Total	0	0	0	0	1	1

View scan details in Wiz

To detect these findings earlier in the dev lifecycle, try using Wiz Code VS Code Extension.

[germa89]

Ignoring bandit failed job because of PyCQA/bandit#1041

See the code:

https://github.com/ansys/pymapdl/blob/3b799a256dd03a8f17e0d520da4857b14a38f9bc/src/ansys/mapdl/core/launcher.py#L2833-L2842

See my log: https://github.com/ansys/pymapdl/actions/runs/11570941634/job/32207760895?pr=3497#step:2:2452

Pinging @ansys/pyansys-core for visibility.

GitHub

pymapdl/src/ansys/mapdl/core/launcher.py at 3b799a256dd03a8f17e0d520da4857b14a38f9bc · ansys/pymapdl

Pythonic interface to MAPDL. Contribute to ansys/pymapdl development by creating an account on GitHub.

GitHub

feat: support for launching an MAPDL instance in an SLURM HPC cluster · ansys/pymapdl@3b799a2

Pythonic interface to MAPDL. Contribute to ansys/pymapdl development by creating an account on GitHub.