Inspired by these:<p>http://news.ycombinator.com/item?id=2475729 (Jedberg on Reddit's recent outage)<p>http://news.ycombinator.com/item?id=2476052 (Skynet, EC2, and Zencoder)<p>Why did Reddit, Zencoder, Heroku, and myriad other services go down, without Netflix encountering the same problem? Can anyone get into the details on what's different about their architectural setup that allowed them to handle the failure gracefully?