> And when some component of the server fails, your app is unavailable until you can repair it. So you have some downtime. Big deal. If this happens once every

> And when some component of the server fails, your app is unavailable until you can repair it. So you have some downtime. Big deal. If this happens once every few years and you need a day to repair it, your uptime is still better than AWS. Not just everyone hosts a realtime API millions of users depend on every second of the day.