Red Hat's Head Of Products On Competing With Docker, Teaming With AWS And A Container Services 'Renaissance' For The Channel

Containers Go Mainstream

At open-source technologies giant Red Hat, executives say it's their company—not Docker—that is the player to beat in containers for the enterprise. Paul Cormier, an executive vice president who heads Red Hat's technology and products organizations, sat down with CRN during Red Hat Summit 2017 in Boston and had plenty to say about the competition around containers. "You need a production-ready environment to really start to deploy [containers] and bet your business on it," Cormier said, "and we're the only one that's proved we can do that."

He added later, "I'm not trashing on Docker specifically. I'm just saying, I haven't seen another container solution from another vendor that I would call a successful commercial Linux vendor."

Following are excerpts of our interview with Cormier.

The majority of Red Hat's product news this week seems to be related to containers. What should we take away from that?

Containers are Linux. That's what you should take away. Container technology has been around for a long time. But with hybrid cloud now emerging, containers are now the killer app for hybrid cloud because they really allow you to take the application to the different types of footprints of the hybrid cloud. But having said that, it's Linux, just carved up in a different way … now that you're starting to see it get into production, and being part of customers' hybrid environments, frankly they're realizing it is Linux. And it needs the same security updates, functional updates, support, life cycle, certifications, all of the things that Linux and RHEL [Red Hat Enterprise Linux] gave in the past. You've been hearing so much about it up until now because it is the killer app for cloud. You're hearing so much about it now from us because it's starting to go into production, and now people need it in not a 'play' version anymore, but in a production version.

As far as production deployments of containers, has the last year been a turning point?

We've had containers embedded in RHEL since RHEL 7, which was three years ago. So the technology has been very solid for a long time. But what's really changed starting now is now you're starting to see OpenShift.io, you're starting to see the developer tools start to center around containers. So it's not necessarily the containers themselves are getting more stable, but you're seeing the tooling around it that's making it much easier to build apps, to support apps, to manage apps. And you're starting to see the ecosystem—which is exactly how RHEL happened—now you're starting to see commercial software vendors look at containers as a way to package and ship their products as well. That's where we announced the Container Health Index—that's aimed at commercial vendors getting their stuff on containers and ready for the enterprise. So I think you're starting to see surrounding things around it getting it more ready for the enterprise. And of course, platforms like OpenShift that now help our customers build, deploy, and manage it as well. There's sort of been a crescendo building up, and now you're hitting production with it.

What is your pitch for why enterprises should go with Red Hat as a container platform?

Because it's Linux, and you need a production-ready environment to really start to deploy it and bet your business on it, and we're the only one that's proved we can do that. Anybody can say they can do it, maybe some others can, but really we're the only ones that've proved in a scalable way that we can bring you Linux-based open source-based technologies to the enterprise. I think we've shown you some of the biggest companies in the world that are running our containers in production.

Does Red Hat have the most production deployments of containers of any vendor?

As far as we can tell. The other guys who think they're container providers, they have to run their own Linux now. You have to be a commercial Linux vendor in order to credibly do containers. Or put the other way, if you're going to do containers, you're going to be a commercial Linux provider. So for example, what Docker introduced last week at DockerCon, Linux Kit. Basically, they've said, 'OK, you've got to be a Linux provider.' Their solution is for the end user, the customer, to build their own kernel. That's not enterprise-ready—that's not how enterprises roll. Enterprises don't want to build their own Linux, and they're the only ones that can support it. So it's those proof points. I'm not trashing on Docker specifically. I'm just saying, I haven't seen another container solution from another vendor that I would call a successful commercial Linux vendor.

What are your thoughts on Docker rebranding the open-source project as Moby?

I think it's confusing for contributors to the project. There are somewhere on the order of 1,600 contributors to the previously known Docker project—many of which are Red Hat employees. And I think it's confusing for the contributors—that have been contributing upstream to a project that they thought was called one thing—to have it overnight unilaterally changed to another thing. But only time will tell on that.

You blogged recently about how the nature of open source means you are often contributing to competitors.

When one of our customers finds a [Linux] problem, and we fix it with our engineers, we give that fix back to the upstream community. So every other distribution can funnel that down into their distribution and use it. We're the only major company that has a 100 percent open-source policy.

You've contributed to the Docker project which Docker Inc. is obviously building upon.

Right. So did 1,649 others. We've all contributed to that upstream project. There's still a lot of unanswered questions. What happens with the name change in the community, what happens with Docker's CE [Community Edition] and EE [Enterprise Edition]? I don't even know if those products will be 100 percent open. If it's an open core model, it's not 100 percent open. We don't think that open core models work very well. We think complete open is the only way—you're either in it, or your not.

So who is going to be the 'Red Hat for containers?'

Red Hat is already the next Red Hat for containers. That is just the case, because it's Linux. The way I put it, we've got a 14-year head start. Because it's Linux, and we've had the commercial model for Linux for 14 years. It was the same thing with OpenStack. It was Linux underneath, and all the other vendors that fell by the wayside, a lot of the reason was they didn't know how to do a commercial Linux model. We weren't very good at it when we started either. Why do we feel so confident about [containers]? Because we've done this before. We've done it for the last 14 years. We've made lots of mistakes along the way, we've learned from those mistakes, probably still have more to make. But you sort of get good at something when you've done it for a long time.

How fast are your cloud products growing?

We had financial analysts' day yesterday, and we typically don't split out revenues by products. But he gave our financial analysts and shareholders a glimpse yesterday, and grouped our major new cloud products—OpenShift, Open Stack and I think CloudForms was in there—and basically showed that at 100 percent growth. RHEL by itself is still low- to mid-double digit growth, but the new cloud products are 100 percent growth. I think that's telling us that people are starting to get it.

I wanted to ask about AWS also. They're a partner of yours, but they're also doing a container platform. So they are a competitor also then?

I don't think we compete with them. I guess, yes and no. Yeah, they're doing their own Linux, yeah, they're doing their own container platform. But what drove this partnership for both of us [is] our customers are asking for a hybrid environment. What that means is they want a common operating environment for the applications across various types of on-premise footprints, all the way out to public clouds. And now, all the way out into multiple public clouds. And that's what we're doing with OpenShift because of RHEL, because RHEL is the piece that gives that consistency for the app. That's what they've been asking for.

Is this a change in approach for AWS?

I think the good news for everyone is that Amazon is starting to recognize that it's a hybrid world. If you're going to run 100 percent of your applications in Amazon, or any other public cloud provider, and take advantage of all those services, because that's the only place you're ever going to run—in a lot of cases, that's not reality. Customers want to be able to run in multiple places, have a common operating environment, so it's more manageable and secure. That's where we come in. So this helps both of us. Because we're able to now give that common environment, all the way across out to Amazon, which we always have—but more importantly we're giving the ability for the application developers to invoke Amazon services even from the OpenShift platform. So they can still keep the consistency of the build environment, the consistency of the container run environment, but call Amazon services if they so choose.

How are OpenShift and containers a good fit for channel partners?

I think it's a renaissance for [channel partners]. Linux and RHEL sort of became a boxed product—you'd buy it in a box, and basically you'd install it in a server and off you go. But what we're doing now with all these technologies, because Linux got so prevalent, all this innovation happened on Linux, you've got all these new complexities that are deploying on these new technologies. So people are now re-architecting their infrastructure, they're re-architecting their app development infrastructure, and they're re-architecting their management infrastructure. They're all going hand-in-hand. So customers need a lot of help moving their environment and implementing these things so they can get into the next generation of computing. So I think that's good for our partners, because now they can take our products and work with customers, and help them get to the hybrid cloud model. We have a very small services organization—our strategy with services is to prime the pump. We want to leave all of that for our partners.