Hardware is always going to fail eventually. Moving to AWS caused NetFlix to write better code to deal with these failures. Failures were always going to happe

Hardware is always going to fail eventually. Moving to AWS caused NetFlix to write better code to deal with these failures. Failures were always going to happen, even in their own datacentre. What they have now is a more fault-tolerant system which should have less downtime overall.