Comment by snickerbockers

How is it even possible for all aspects of such a massive enterprise to all share a single point of failure like that? And why can't they just cut their losses on the past N days of business, restore all these servers from snapshots and get back to business?