Comparison
Contributing factorvsRoot cause
Contributing factor
the deploy triggered it, and the missing alert, the stale runbook and the one-person rotation all had to be true as well.
One of several conditions that had to hold for a failure to happen or to last as long as it did. It is the replacement for root cause in most modern incident practice, because a list of factors gives you a list of places to intervene rather than one thing to blame. It also matches how the failure actually felt to the people in it, which is what makes the review honest.
Full entry →Root cause
the report named one cause, and three of the people who were there could each name a different one.
The single underlying reason a failure occurred — a framing borrowed from mechanical systems that fits software poorly. Complex systems fail through combinations, so the search for one cause usually stops at the most recent human action, which is the least useful place to stop. Reviews that ask what conditions made the failure possible produce better actions than ones that ask what the root cause was.
Full entry →