Share The Learning

A lot of engineers move on immediately after landing a fix, but the best ones take the time to reflect and prevent similar issues in the future. Here are the core points from the lesson:

• Retrospectives help you identify root causes and system weaknesses so you can either prevent a future incident or make it much easier to detect and fix
• A good postmortem analyzes what went well and what didn’t across detection, notification, response, mitigation, and analysis, with detailed documentation to capture all learnings
• Writing a clean, detailed doc with screenshots and explanations makes it much easier for future engineers to learn from the incident without having to dig through chaotic logs or chat threads
• Big tech cultures like Facebook normalize failure and focus on minimizing damage and learning quickly, often using formal SEV reviews to improve systems and processes over time