Arepo Production Run

1. Finalize production configuration

• Lock Config.sh and runtime parameter files.
• Document the exact source revision and version-control it.
• Compile a dedicated production executable.

2. Stage data and scheduler job

mkdir -p /scratch/<project>/arepo-prod/{ic,output,logs,restart}
cp <ic-file> /scratch/<project>/arepo-prod/ic/

Example SLURM launcher:

#!/bin/bash
#SBATCH -J arepo_prod
#SBATCH -N 8
#SBATCH --ntasks-per-node=64
#SBATCH -t 24:00:00
#SBATCH -o logs/slurm-%j.out

cd /scratch/<project>/arepo-prod
srun /path/to/Arepo params_prod.txt

3. Operational safety

• Run a short pilot job with production parameters first.
• Monitor timestep evolution, memory usage, and I/O throughput.
• Verify that the restart workflow works before submitting long jobs.

4. Post-run quality assurance

• Check global conservation quantities and expected scaling behavior.
• Compare key diagnostics against pilot and baseline runs.
• Archive the executable, parameter files, scheduler scripts, and all outputs.