Situation
Production conversations kept assuming container and orchestration patterns I would not roll onto a customer site cold.
Task
Build a personal lab that could fail safely so production advice was earned, not guessed.
Action
Stood up home-lab Docker and Kubernetes paths, broke things on purpose, and documented what actually survived reboot, upgrade, and bad config.
Result
Later production guidance and support calls rested on scars from the lab, not slide-deck confidence. (Stub — expand with concrete stack and incidents.)