Kate Matsudaira interviewed at Velocity 2011
summary
Launch failures fall into three concrete causes: self-inflicted deployments that break systems, environmental problems like flaky networks or third‑party APIs, and load/resource exhaustion (traffic, DB connections). Immediate response emphasizes clear communication, appointing a leader, triage to fix the highest‑impact issue first, and choosing a temporary fix versus a long‑term one. Longer term requires regular postmortem reviews (not just after disasters), no personal blame, published notes with action items, monitoring, unified logs, offline customer contacts, and a rehearsed “when it fails” plan.
AI-written summary · about
related talks
06mIgnite: Failure and the Third Way
26mMission Critical Mindset Building a Culture of Resilience in DevOps Teams
41m‘Failure’ as Success: The Mindset, the Methods, and the Landmines
30mLearning From Failure: Post Mortems
50mDevOps Project Failures - Lessons Learned and Strategies for Recovery
41mYour Systems Aren't Stuck - Your Teams Are
28mThe Failing CTO - What we can learn from others' mistakes
20mDoing Things the Hard Way