Reproducible, disposable, observable

Three habits that show up in every corner of the lab:

• Reproducible. Every VM is a clone of one golden image, configured by cloud-init. Even these docs were built through an API. The goal is that nothing important exists only as a thing someone once clicked into being.
• Disposable. Because creation is automated and teardown is clean (snippets detached, temp secrets removed), VMs are cattle, not pets. You can destroy and rebuild without ceremony — which is also what makes you brave enough to experiment.
• Observable. The logging agent is baked into the template, so every machine ships logs from its first boot. Metrics watch the host. You see problems on a graph before they become outages.

The mindset: build things you can rebuild, throw away, and see into. A system that's reproducible isn't fragile; one that's disposable invites experimentation; one that's observable doesn't surprise you. These three turn "infrastructure" from a museum of hand-crafted artifacts into something you actually operate.