Rewrite CI orchestration into helm charts

Initial work

• Simplify the scripts from simpler to more complicated cases
  • kdc key propagation removal (common, not CTA specific)
  • look at serialized/synchronized parts of the create_instance.sh script: dependency chart? secrets VS standard configmaps?
• helmifying hurdles?
  • separate helm chart for the eos side?
    • simplify ctageneric as it just has to run cta- processes?
    • dcache helm chart integration for the storage side?
• Evaluate helm uses outside of the pure CI use case
  • developer killing ctafrontend pod to upgrade it to a newer version (k8 deployment)
  • Test against a newer version of EOS? Coming from EOS official repo.
  • A dev version of EOS compiled in minikube_CI context?

Presentation of achievements

2024-08-22 15:00 in the cta developer meeting slot.