Hacker News new | past | comments | ask | show | jobs | submit login

Looking at the status graphs, it seems like there was some clearly anomalous data starting around midnight, about 9 hours before the actual outage "began". Maybe a gradual botnet ramp-up, and 9:27 AM is when it got bad enough to overload some critical service? (Or really any other threshold-based failure scenario.)



or a bad commit being deployed on their fleet.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: