It's possible they deal with so much load that they considered a day's worth of traffic to be sufficient load testing: > The system had worked well with stream

It's possible they deal with so much load that they considered a day's worth of traffic to be sufficient load testing: > The system had worked well with streaming at this level for a day before the incident started, so it wasn’t initially clear why it’s performance had changed. And a short note later on how much load their caching system sees: > These databases were unaffected by the outage, but the caching system, which regularly handles 1B requests-per-second across its multiple layers during regular system operation, was unhealthy.