Postmortem Action Items: Plan the Work and Work the Plan
summary
Postmortem practice centers on concrete, measurable follow through: classify action items as prevention, mitigation, detection, require at least one P0 or P1 bug tied to root cause, and set SLOs so high‑risk fixes don’t languish. Automate where humans fail — a config change that auto‑reroutes traffic removes a recurring maintenance outage for six years. Drafts aim for 24 business hours, and the single biggest operational win is switching from 100% rollouts to phased rollouts (1%, 5%, 20%) to limit blast radius.
AI-written summary · about
chapters (20)
- 0:00Intro
- 0:47About the speaker
- 1:09Postmortems
- 1:53Action Items
- 3:10Antipatterns
- 6:20Balance
- 6:38Symptoms
- 7:25Diagnosis
- 9:19Humans as a root cause
- 13:11Not going beyond prevention
- 15:39Best practices
- 15:58Prioritize and classify
- 17:06Executive focus
- 18:18Ben Traynor
- 20:05Postmortem Reviews
- 21:49Reports and Dashboards
- 22:30SLOs
- 23:05Closing Bugs
- 24:47Summary
- 26:15Resources
related talks
10mSimplify your postmortems and focus on scaling
28mBlameless Postmortem Culture
12mUsing postmortems to break out of infinite loops
45mIncident Management - Talk the Talk, Walk the Walk
22mAfter the dust has settled: learning from incidents
23mYou put out the fire on production, what’s next?
15mPragmatic Incident Response: Lessons Learned from Failures
15mMaintaining reliable systems