I mapped out how debugging works during production incidents (nemorize.com)

This roadmap focuses on:

triage before diagnosis

when dashboards lie

why doing nothing is sometimes correct

partial failures and cascading effects

humans under stress

turning incidents into better architecture