A lot of smart people discover their fully redundant architecture has a single point of failure during an outage. AWS has had a fair number of large profile ou

A lot of smart people discover their fully redundant architecture has a single point of failure during an outage. AWS has had a fair number of large profile outages and service degradation, including a few that crossed zones, even though that "can't happen".