Avengers Assemble – The Thanos Incident
summary
A retrospective reconstructs the timeline leading to the Snap, names the contributing factors instead of hunting a single root cause, and shows concrete practices: use PagerDuty style paging and clear on‑call rotations, appoint a single incident commander, run a blameless post-mortem that asks how and what not who or why, call in cross‑team experts, document decisions and data, and count the surprises (time travel aside) to turn the Avengers’ chaotic response into repeatable incident response improvements.
AI-written summary · about
chapters (11)
- 0:00Start
- 9:10THERE IS NO SINGLE ROOT CAUSE OF MAJOR FAILURE IN COMPLEX SYSTEMS, BUT A COMBINATION OF CONTRIBUTING FACTORS THAT TOGETHER LEAD TO FAILURE
- 15:49FUNDAMENTAL ATTRIBUTION ERROR
- 17:02CONFIRMATION BIAS
- 18:27HINDSIGHT BIAS
- 19:35NEGATIVITY BIAS
- 23:19ABSTRACT TO AN INSPECIFIC RESPONDER
- 24:56CONTRAST WHAT YOU DID NOT INTEND WITH WHAT YOU DO INTEND
- 29:30DELEGATE AND COORDINATE
- 29:53DECISION MAKER
- 30:09SINGLE SOURCE OF TRUTH
related talks
37mIncidents are a new normal
15mIncident Ready: Chaos Engineer Your Incident Response
39mNavigating the Chaos: A Holistic Approach to Incident Management
27mLessons from Incident Management and Postmortems at Atlassian
20mData Incidents - When things go wrong
34mIncidents are new normal
29mComposing the Response: Building an Incident Pipeline from Scratch
27mWe Stopped Thanos from Snapping $100K from Infra