Currently PyFluent session using slurm launcher can get stuck in some systems while running large number of iterations. This is due to capturing of Fluent's stdout which is required for slurm launcher as we read the slurm job id from Fluent's stdout. Similar issue was fixed for the standalone launcher by disabling the stdout capture by default (#1628).

At present, a workaround for the slurm launcher is to use 2 PyFluent sessions. The first PyFluent session will just launch Fluent using slurm launcher and exits. It will use launch_fluent with cleanup_on_exit=False to keep the Fluent process running even after PyFluent exits. The second PyFluent session will connect to that Fluent process and run the long iteration. It'll need to pass cleanup_on_exit=True to connect_to_fluent to exit the Fluent process . Note that, the first session captures stdout, but the second session doesn't which fixes the hang during iteration.

A PyFluent side solution can be developed by implementing the above workaround.