The infrastructure has been rock-solid. I’ve never seen a service grow this fast without any noticeable outages. The architecture and execution are obviously informed by serious experience at Twitter, but it’s also clear that management is giving them everything they need to do the job right.
replies(4):