Red Hat CEO Jim Whitehurst On How He Plans To Win The Container Market
Red Hat's Strategy To Dominate Containers
Red Hat's goal for its OpenShift container management platform is nothing short of total domination of one of the tech industry's hottest markets, according to the company's CEO, Jim Whitehurst.
At the Red Hat Summit in San Francisco, Whitehurst sat down with CRN to discuss his team's laser-focus on positioning OpenShift, a container-native Platform-as-a-Service, as a flagship product that will drive the Raleigh, N.C.-headquartered company to new heights in the coming years.
That container strategy looks to capitalize on the market position of Red Hat Enterprise Linux, and the recent acquisition of CoreOS, Whitehurst said.
The stability and security of Red Hat's ubiquitous operating system, when coupled with the ease-of-use CoreOS has built into its Linux and Kubernetes platforms, will propel OpenShift to the front of the container pack, Whitehurst said.
Is Red Hat determined to lead the container market?
We are laser-focused on making sure we win the container platform.
Containers are Linux. It is. Kernel space is in the lower-level substrate and the user space is in the container. Now that requires something like Kubernetes to orchestrate it, but it is just a different way to deliver Linux.
You don’t have RHEL [Red Hat Enterprise Linux] and a different container platform. That gives us a huge head start because we're pretty good at Linux. But it also is kind of a must-win for us because you can't have a competing container platform in RHEL. It is RHEL in a different manifestation, now with a lot more value because the platform does all the manageability and other stuff.
So for us it's existential to win.
Can you see Red Hat ultimately being to containers what it is to Linux?
Yes. Because containers are Linux. It's almost a tautology.
So the beauty of RHEL, and our derivatives like CentOS, every piece of hardware certifies to it, every application certifies to it. So now you can know you can switch out a Lenovo server for a Dell server and your application is going to keep running.
That's true in OpenStack and the reason we're so invested in OpenStack. It's the same thing. It's where hardware touches software. You kind of get that in the container platform.
If you're the ISV writing an application, do you want to have to write to 10 different container platforms? Probably not. Because it's based on RHEL, every piece of hardware can already run OpenShift because if RHEL runs there, OpenShift runs there.
Why was your CoreOS acquisition so important for the container strategy?
We announced [at Red Hat Summit] a whole bunch of functionality from CoreOS integrating into OpenShift.
What's exciting to me about that, Red Hat is an enterprise software company, so we think a lot about software, and I would argue that OpenShift is by far the best content in a container platform. Super secure, reliable, life-cycled, all the things that Red Hat does so well.
CoreOS took a very different approach. They were operators. So they start off with making this next-generation application platform really easy to use, with automated operations, over-the-air updates, all that kind of stuff.
What's the advantage of bringing together Red Hat and CoreOS technologies?
The RHEL bits—the 10-year lifecycle, never-break binary compatibility, super secure, common criteria certifications—all the things we do to make RHEL what it is, we do it across all of our products. What we're bringing to the table is this really well-life-cycled, rock-solid content.
But we don't necessarily think of ease-of-use or an operator's perspective quite the same way [as CoreOS], so seeing what they've done in terms of an operator's perspective, how you make this drop-dead simple, you combine the two and its pretty slick.
How will CoreOS's Linux and Kubernetes platforms integrate into Red Hat products?
Tectonic is their Kubernetes platform, and OpenShift is ours. We're basically taking all their set of features, Quay, and Operators, and their update mechanism, and integrating those into the next release of OpenShift.
We're calling the operating system RHEL CoreOS. It uses their paradigm around updating. That's the Linux in the Kubernetes platform, the kernel space, the lower level. We're basically having RHEL content with their paradigm for how it updates and works.
They have a bunch of stuff they put on top which we're now putting on top, underneath their paradigm on how you should update an OS. We are adopting that but it's with the RHEL content, so it really is bringing the best of the two together.
Why should competitors like Docker and Pivotal be worried about Red Hat?
As you can imagine, beyond us, everyone wants to say Linux doesn't matter. But it is Linux. It's just Linux delivered in two pieces.
Because we're the biggest Linux player, if you're trying to compete with us, you don’t want to talk about Linux because you'll lose to us.
And so I do think Docker, I can't remember what Linux they're putting in it, or PKS [Pivotal, Google, VMware collaboration] with Photon [lightweight Linux distro from VMware], or [Pivotal] Cloud Foundry with some bastardized Ubuntu OEM thing, honestly, I think we have a big head start [over them] because we know how to life-cycle an operating system really, really well.
What do you see as Red Hat differentiators around security?
Recognize the entire user space of the operating system is in the container.
So you go to an ISV and say, hey, you're going to deliver your applications as containers, great. Well, you now have the user space of the OS in there, and guess what, 90-something percent of the vulnerabilities in the operating system happen in the user space.
Are you going to life-cycle? Regular ISVs don’t want to be in that business. Are you going to patch it every time there's a security vulnerability? By the way that happens a couple times a week. So all of a sudden, our ability to go to ISVs and say, maybe you want to make those RHEL containers, and we'll life-cycle the RHEL piece in there, it starts to be a compelling value-proposition.
Do you see customers considering those security issues when choosing platforms?
We've had customers who have used competing platforms to get all the way to the CSO, and the CSO is like, "That ain't certified here, you're not bringing that in there." They don't have any type of Common Criteria Certifications. They just can't pass the security audit processes and all the other stuff, and then they come back and go, I think we'll go with OpenShift.
That's the big advantage we have.
Why is Red Hat's standing in the open-source community important?
It's more complicated than this, but a container platform is primarily Linux and Kubernetes. And we're the second-largest contributor to Kubernetes—we're neck-and-neck with Google now after acquiring CoreOS. We're the largest contributor, or neck-and-neck with Intel in Linux.
So we can drive those roadmaps forward.
What's the significance of the deal you announced with IBM this week?
IBM putting all their software on OpenShift, I think that shows they believe it can run mission-critical, and of course it can because it's RHEL.
But technically, you can run WebSphere on [Microsoft] Azure if you want to. Wherever OpenShift can run, now WebSphere can run. That's value to IBM to say, now you can run our software anywhere because OpenShift runs anywhere, and they've certified to OpenShift.
Are public cloud container services a competitive threat to Red Hat?
You can run OpenShift on all the clouds. Our value proposition is: write to OpenShift, you can run it anywhere. We have a service broker concept, so if you wanted to run OpenShift on Azure, but you want to call Amazon's service, have at it. You can do that.
Our model is: buy our infrastructure and you can run it anywhere. Where I think Amazon service or Google service is: come by our cloud cycles and we'll give you the container platform.
We announced a big thing with Microsoft but they have their own container platform. We work a ton with Amazon [around] OpenShift, but they have their own container platform. Yeah, they compete, but it’s a different way to compete. It's also complimentary.
If you want hybrid, [OpenShift] is the only place to go.