[Music] my name is Yuri Israel FSK and I'm BPO of cloud and platform engineering Netflix is the world's premier service for enjoying movies and TV shows to be successful and out-compete the other players in the market we have to be great in a number of things we have to have great content personalization device penetration great customer service UI and so forth and operating data centers is not one of those things in fact it's a distraction for us the original motivation for Netflix moving into the cloud came from a database corruption that we experienced in the
August of 2008 and for a few days we have problem shipping discs tour DVD customers and at that time we realized that going forward what we really needed was a more full tolerant system that wouldn't be privy to failures like this AWS was at that time as it is right now a clear leader in the cloud space and you know had the vast capacity that we would actually grow into in 2009 we migrated a lot of our transcoding then that was also a very positive experience and then as far as the streaming applications that migration
really happened in 2010 in the summer of 2010 we launched the iPhone streaming service all over 800% out of cloud we wanted to become experts at using the cloud so we developed a number of tools to help us be better at that we constantly have to test and verify your ability to survive any kind of failure and that's the reason why we developed tools like the chaos monkey which actually goes and intentionally breaks things in production just to make sure that we can survive those kinds of failures so scalability especially for streaming service like Netflix
which has been growing tremendously over the past few years is really critical it's important for us to be able to add thousands of servers and terabytes of storage at minutes notice in the past one we've had really high priority critical encoding tasks in front of us we would sometimes spin up many thousands of servers and complete the tasks within a matter of hours or a few days and then spin a down laughter we no longer need it we've actually increased our level of reliability by more than an order of magnitude reducing the complexity of running
in both the data center in the cloud was a big part of that but using redundant services like multiple availability zones has helped us achieve that higher level of availability so we get both the benefits of higher efficiency and higher availability also for Netflix which is an international company it's been really helpful for us to leverage the global regions that AWS provides so wherever you happen to be you can access the service and enjoy the benefits of the Netflix product [Music] you