[Documentation] Explain advanced scheduling features in flux

[jan-janssen]

Example script used in the exascale project:

#!/bin/bash
#SBATCH --qos=debug
#SBATCH --time=02:00:00
#SBATCH --chdir=/ccs/home/janssen/pyiron/projects/2023-09-24-dftfe-current
#SBATCH --get-user-env=L
#SBATCH --nodes=100

srun --cpus-per-task=56 flux start -o,--config-path=./ python test_queue_v10.py

And then in the working directory, we place the sched-fluxion-qmanager.toml file to configure the scheduling behavior of flux:

[sched-fluxion-resource]
match-policy = "lonode"

More details about the different scheduling policies are available in the flux documentation:
https://flux-framework.readthedocs.io/projects/flux-sched/en/latest/man5/flux-config-sched-fluxion-resource.html#resource-match-policies