Who to blame for all your problems
summary
Argues that postmortems should be blameless by default and that the team assume correct decision given available info so human error is not a terminal conclusion and honesty increases. Prescribes concrete process steps: draft the postmortem within a day, assemble a timeline with timestamps and Bash history, run the meeting led by someone other than the incident lead, list plural root causes, and convert findings into SMART action items such as pin docker and docker-compose in Ansible and turn on instance termination after four failed health checks to prevent recurrence.
AI-written summary · about
chapters (15)
- 0:00Intro
- 0:49What is a postmortem?
- 2:37Why blameless postmortems?
- 5:24Anonymity !Blamelessness
- 6:40Accountability Without Blame
- 9:58National Transportation Safety Board
- 10:48Getting started
- 11:14Postmortem parameters
- 11:58General Summary
- 14:01What we learned
- 15:23Before the meeting
- 16:07Running the meeting
- 18:20Context Error: Alternative Actions
- 18:40Context Error: Cherry Picking
- 19:16Context Error: Shopping Bag
related talks
28mBlameless Postmortem Culture
31mThe blameless post mortem: how embracing failure makes us better
16mInterview
05mWho's to Blame?
43mPostmortems, Learning & Blameless Culture
51mEmbracing Failure and Learning from Our Mistakes with Effective Post Mortems
31mPost-mortems: Building better software together.
28mPost-mortems or How to Walk Away Happy from a Carnage