The biggest news to hit the data centre universe lately is surely that Netflix, after seven long years, is finally shutting down the last of its data centres in a bid to host nearly its entire offering in the cloud (Amazon Web Services, specifically).
[easy-tweet tweet=”Why has it taken 7 years for @Netflix to move wholly to #AWS?” user=”comparethecloud”]
The move has made the headlines big time, positioning Netflix as a cloud pioneer and AWS as the Holy Grail, but the real question that sprang to mind when the headlines emerged was this: Was it really such a surprise? Furthermore, why did it take seven years to make this move?
The Cloud: Otherwise Known as Someone Else’s Datacentres
The headlines are heralding the approaching “shutting down of the final Netflix datacentre,” as though it’s a monumental move, which should really prompt us to ask: why is one of the world’s largest home streaming video businesses opting to offer its services via one of the world’s largest public cloud providers such a big deal?
why is one of the world’s largest home streaming video businesses opting to offer its services via one of the world’s largest public cloud providers such a big deal?
Netflix is only doing what other companies have been doing in droves for several years: moving to the cloud. And while “the cloud” may seem the solution to an array of problems presented by aging, expensive and underpowered datacentres, when we refer to “the cloud,” what we’re really talking about here is just another collection (albeit much bigger and fully serviced) of datacentres. In this case, it just happens to be owned and run by Amazon.
But Why did it Take Seven Years?
Seven years ago Netflix had only been in the video streaming business for two years and the demand these services was not near high enough to consider branching out beyond its existing infrastructure.
However, seven years on Netflix Vice President of Cloud and Platform Engineering, Yury Izrailevsky, had some great insight on the journey of this migration, which is possibly some of the most interesting behind the scenes migration information we’ve seen outside our own migration projects:
[easy-tweet tweet=”Moving to the #cloud was a lot of hard work, and we had to make a number of difficult choices along the way says @Netflix”]
“Given the obvious benefits of the cloud, why did it take us a full seven years to complete the migration? The truth is, moving to the cloud was a lot of hard work, and we had to make a number of difficult choices along the way. Arguably, the easiest way to move to the cloud is to forklift all of the systems, unchanged, out of the data center and drop them in AWS. But in doing so, you end up moving all the problems and limitations of the data center along with it. Instead, we chose the cloud-native approach, rebuilding virtually all of our technology and fundamentally changing the way we operate the company. Architecturally, we migrated from a monolithic app to hundreds of micro-services, and denormalized our data model, using NoSQL databases. Budget approvals, centralized release coordination and multi-week hardware provisioning cycles made way to continuous delivery, engineering teams making independent decisions using self service tools in a loosely coupled DevOps environment, helping accelerate innovation. Many new systems had to be built, and new skills learned. It took time and effort to transform Netflix into a cloud-native company, but it put us in a much better position to continue to grow and become a global TV network.”
Perhaps the more relevant question is this: how long would it take a similar sized provider to move its services to the cloud starting in 2016? And, perhaps more importantly, would they do it in the same way Netflix has?
Not Your Typical Migration Story
The Netflix migration story is an interesting account and one any technologist would like to hear more about, but it’s not a typical migration story. It’s a great testament to Netflix that they were able to actively transform themselves this way, building a new infrastructure and achieving continuous delivery on so many levels while at the same time scaling and moving piece by piece to the cloud.
It’s a great testament to Netflix that they were able to actively transform themselves this way
The ability for a company to fund a re-design of their entire application landscape is unusual, and strategically is the right thing to do – as Gartner talks more about datacentres of micro-services in the same way.
However, in comparison to other large enterprises that have been around some time, Netflix is very new, and so to re-architect everything would be relatively simple as the company wouldn’t have any significant legacy apps.
More established companies are not looking to downsize their datacentre footprints as such because in the majority of cases, it’s unlikely such a move is needed. Most large enterprises are happy with the solution they already have, which is often a hybrid cloud infrastructure that makes best use of all worlds. The cost of moving entirely to the cloud could well be astronomical, which makes it all the more unappealing.
Beyond this, most companies that are older than seven years’ old have legacy applications that are simply not cloud-friendly. Such organisations will leverage the commodity nature of cloud providers (burst and retraction of compute and network capacity as two examples) for specific applications, rather than shoe-horn every system into a service model that wouldn’t be appropriate for a proportion of their systems.
[easy-tweet tweet=”Most companies wouldn’t look to migrate wholly onto the #cloud because of significant legacy apps”]
Will it take seven years?
No, because unlike Netflix, most companies are selective about what they migrate to the cloud, rightly recognising that the cloud is not suitable for every application that they host within their datacentre.
The best approach would be not to fork-lift applications directly into the cloud, but to select those that are easy to move and provide best value. Planning for this to take months, rather than years should be the expectation. If you’re an established organisation, moving everything to cloud is sub-optimal in the same way that holding everything in-house is also not the best application and value strategy.