The public Infrastructure-as-a-Service (IaaS) cloud is big and set to get much, much bigger, according to IDC. How big? $43.9 billion by 2020, up from $12.5 billion in 2015.
Not surprisingly, Amazon Web Services (AWS), which today accounts for most of that IaaS revenue, is expected to continue to dominate the IaaS market, “as economies of scale and continued investment drive the cycle of adoption and growth.”
OpenStack, for all its community bluster, isn’t slowing the AWS train. OpenStack appeals to enterprises that remain skittish about public cloud and want to pretend to be getting the same benefits.
Container management project Kubernetes, however, may be positioning the “Anyone But Amazon” crowd to mount a more credible threat to AWS, according to Redmonk analyst James Governor.
Kubernetes and AWS
Containers are hot with developers, as they increase productivity, security, and offer a host of other benefits. To date, however, they’ve been hard to manage at scale. Enter Kubernetes.
Kubernetes, hatched at and open sourced by Google, is arguably the strongest community in the container orchestration space. Such community, however, may have little to do with holding hands and singing kumbaya and everything to do with killing AWS, as Governor deftly describes:
Microsoft, Red Hat, and Google Cloud Platform are all now aligned around Kubernetes. While at first glance this new alliance of strange bedfellows might seem to be a response to the rise of Docker and the Docker Pattern–and indeed there is no doubt the enthusiastic growth in Kubernetes is partly driven by concern that Docker will own too much real estate of the new infrastructure world–I believe the overarching threat is Amazon.
At first glance, there’s no reason to think Kubernetes and AWS can’t coexist (as indeed they can and will at many organizations). But, reading through Erez Rabih’s comparison of Kubernetes to Amazon’s Elastic Container Service (ECS), you start to get a sense of what’s at stake.
“As many companies are starting to use Docker as their main infrastructure and delivery mechanism, orchestration frameworks become the heart of the system and affects the way we develop, ship, run and update software,” Rabih said.
OpenStack purports to give companies a way to build rival clouds, but mostly it has simply served as a way for enterprises to pretend to turn their private data centers into “clouds,” without getting the true benefits a public cloud provides (in terms of agility, among other benefits).
But what about cloud?
But Kubernetes, and the containers it manages, is something more fundamental. It’s an alternative way to think about infrastructure and delivery.
It’s not that Kubernetes somehow obviates cloud, public or otherwise. It doesn’t, and Kubernetes will happily run on AWS and other clouds.
SEE Hey ops teams, developers want control of the data center (TechRepublic)
This, however, is the point. Kubernetes potentially up-ends the idea of running everything in one particular cloud. As Rabih contends: “Kubernetes is cloud agnostic. You can run your cluster on AWS, Google Cloud, Microsoft’s Azure, Rackspace etc, and it should run more or less the same.” There are significant benefits to keeping all your cloudy assets in one place, but for companies wanting to hedge against Amazon lock-in, Kubernetes will be a rational choice.
Of course, all this meta decision-making wouldn’t matter much if the only one supporting it were the CIO, but she’s not. Developers love containers, and developers (and the ops teams that support them) love Kubernetes. This love-fest likely won’t stop AWS’ momentum, but it does offer a compelling, rivalrous way of thinking about enterprise infrastructure, one that many enterprises may elect to embrace.