In the old world, the operating system was the center of the computing universe. In today’s modern application age, container platforms fill the space once occupied by the OS. That fact helps to explain why so much cash and code is pouring into Kubernetes.
What it doesn’t explain, however, is why comparatively few companies are investing heavily enough in Kubernetes. Given the central role Kubernetes seems destined to play in shaping enterprise infrastructure, we should see more competition to contribute, given that more code tends to equal more influence in open source. Oddly, only Google and Red Hat seem to be playing to win in Kubernetes, based on contribution counts.
Coding Kubernetes for cash and influence
Google tops the Kubernetes charts, which isn’t too surprising. After all, Google, in a master stroke of open source brilliance, first released the project as a way to give outsiders access to its internal container management smarts. Google isn’t a charitable foundation, however, and this act helped to position its cloud as the premier place to run Kubernetes workloads. “Build with containers on your laptop,” was Google’s call to developers, “but scale them on Google Cloud Platform.”
But not if Red Hat gets there first.
Call it dumb luck or shrewd strategy, but Red Hat has rocketed up the Kubernetes contributor charts over the past few years, determined not to miss out on “the new enterprise Linux,” as Red Hat’s Daniel Riek has styled it.
Important for the open source leader, Red Hat CEO Jim Whitehurst told me that, whereas in the old enterprise computing world the company could grab roughly 20 percent of the Linux market and still miss out on the 50 percent of the market that stuck with Windows, “containers are Linux.” That is, for a company with Linux in its DNA, the entire market is now part of its near-term, total addressable market.
That’s huge. And it’s why Red Hat—perhaps more than most—saw the need to contribute early and often to Kubernetes, which is rapidly becoming the industry’s operating system for the next few decades.
The stakes are huge. So is the cash. For a company like Red Hat, Kubernetes is a deal-multiplier. As Whitehurst explained, there’s a lot more value in a platform where app services need to run and, because there’s more value, Red Hat charges more for it. How much more? To run Red Hat Enterprise Linux (RHEL) on two sockets sets you back roughly $800. Those same two sockets running OpenShift, Red Hat’s Kubernetes platform? $15,000.
That’s right. Same “operating system” approach, but 20 times as much total addressable market potential. In the last quarter, Red Hat reported a 75 percent increase in OpenShift customers. Kubernetes is the gift that keeps on giving to Red Hat.
The cloud hordes have arrived
Others have been slow to see the opportunity that Red Hat clearly sees, but that doesn’t mean that Red Hat is alone. Aside from Google, the major cloud vendors have been slow to tap into Kubernetes. The most obvious contender, Amazon Web Services, is completely missing out on the industry’s early embrace of the community container platform darling. Whitehurst told analysts that Red Hat doesn’t “often compete directly head to head with the cloud providers,” and that’s likely because 1) most cloud activity happens on AWS, which was MIA in Kubernetes, and 2) Google has the best Kubernetes story of the major cloud vendors but hasn’t been a strong enterprise player.
This has left the ground to Red Hat.
AWS finally released its Kubernetes service in June 2018, and it will almost certainly see a spike in adoption. But maybe not as big as you might expect. It’s right to expect many enterprises to go native on the cloud from the start, using Kubernetes there rather than on-premises (though current Cloud Native Computing Foundation data suggests that many still run Kubernetes on-premises).
Lost in that analysis, however, is the reality that in open source land, code contributors tend to reap what they sow. It’s one thing to offer a Kubernetes cloud service. It’s quite another to give customers confidence that they can influence the Kubernetes roadmap, which currently only Google and Red Hat can do with confidence (and Microsoft to a lesser extent).
Only Google and Red Hat are driving the Kubernetes roadmap, Whitehurst told me, which gives big enterprises comfort that these vendors can support them long-term. (Though, again, Google has not been an enterprise player historically, so it’s unlikely it’s having the same level of strategic discussions that Red Hat does.
For Red Hat, this could result in “tens of billions in revenue,” said Whitehurst. And Red Hat doesn’t have to win all of the Kubernetes market. Again, in the old world of operating systems, the market was split between Linux and Windows, with Red Hat getting just a fraction of the overall market. That sizable share of the Linux market, but fractional share of the overall market, is worth billions to Red Hat today. Anything like the same percentage of the burgeoning Kubernetes market, however, has potentially 20 times the revenue.
Which is why Red Hat keeps contributing code to Kubernetes, and why every other vendor that wants a big stake in its future, and the code that comes with it, should be investing just as heavily.