What's actually broken about post-mortems at your company?

10d • DevOps

What was the most broken part of your post-mortem process? Not the incident itself, the aftermath.For me, the worst part is always the "How did we miss this in staging?" question. It's never a simple answer, and trying to explain environmental drift or non-deterministic race conditions to a VP who just wants a "yes/no" feels like a losing battle. I end up writing a doc that's half technical narrative, half political damage control, and neither half is actually useful the next time something breaks. Curious whether this is universal or just a me problem. Maybe your team has actually figured this out. I genuinely want to know if anyone has a process that doesn't feel like reconstruction work after the fact.

1 comment