How we model checks, runbooks and incidents — and why we made those calls.
Why we model monitoring as named, scheduled runbooks instead of a wall of graphs.