Cloning in the batch jobs?

Do you want to keep the cloning in the batch jobs? Talking mostly about ILDConfig now, but these can become quite large and cloning can take quite a bit. If it's at all possible to lift the cloning to a "general setup step" that happens before the submission, etc. happens, I would recommend doing that. The obvious disadvantage is that it becomes user responsibility to make sure that things are consistent and that the config that was used for submission is reproducible.

Potentially something for another PR.

_Originally posted by @tmadlener in https://github.com/key4hep/k4DetPerformance/pull/3#discussion_r1623916895_

Start by tracing where ILDConfig is cloned during batch-job submission and how the general setup step is organized. Clarify whether cloning can happen before submission, then define how consistency and reproducibility of the submitted configuration will be preserved. Done means the cloning responsibility and configuration flow are explicitly resolved.

Written by the indexing model from the issue text.