Labs

Reproducible environments that break on purpose. A case study explains a failure; a lab lets you cause it, watch the signals move, and practise the investigation before you have to do it at 03:00.

Each lab is intended to contain:

text

README.md          the scenario, and what "solved" looks like
setup/             scripts or manifests that build the environment
scripts/           the fault injection
expected-output/   what the signals should look like when it breaks
solution/          the walkthrough, kept separate so you can try first

Planned

Lab	Pairs with
`kubernetes-crashloop`	CrashLoopBackOff After a Release
`container-memory-limits`	OOMKilled Under Load
`prometheus-alerting`	From alert fatigue to SLO-based alerting
`terraform-drift`	Drift detection
`nginx-troubleshooting`	Intermittent 502s