Invest in the contributing system instead of ending the investigation at the person who made the last change.
What to look for
- 01
Connect the immediate fix with the conditions that allowed the failure.
- 02
Make improvements at more than one level of the system.
- 03
Use deployment feedback to detect and contain a bad change.
Keep in mind
A chain of “why” questions can oversimplify an incident. Preserve branching causes and uncertainty rather than forcing a single root cause.
Try this next
Draw two independent paths that contributed to a recent incident. Add a detection or containment improvement on each path.
Get the reliability templateNotes and exercises by Awesome CTO. The original work belongs to Eric Ries.
