kheops/mkdocs/docs/guide/best_practices.md

• 104: Best practices
  • Going into IaC
    • Use git to track your tree
    • D.R.Y.
      • Things are quite easily moving
  • Code structure
    • Profile and Class
      • Role and profiles key lookup
    • Hierarchies
      • Foreman
    • Use environnements, site, locatation ...
    • Common keys vs dedicated keys
      • The profile key, default placeholder
    • The resource modele
      • It's like a puppet resource, a catalog of items to apply
      • It's possible to use this model with ansible, and it change radically the way Ansible can be used then. See integration.
  • Debugging
    • Use the explain mode
      • And trace mode
    • Use GNU tools
      • tree
      • grep -Rw .
      • git status -sb
  • With ansible.
    • Use a strict naming scheme, apply your Ansible usual name schema into Kheops
    • You can put some jinja variable into Kheops, Ansible will be able to replace and understand them during the runtime
    • Dynamic inventories and ENC
    • Apply products
    • Apply roles/profiles pattern
    • Use resource based roles => See my collection, it just works