Red Hat was founded just two years after Linux was announced. Since then, company has continued to evolve and stay a leader in many areas — and with everything fully open source? What your secret sauce?
Part of our secret sauce is that we are open and we’re passionate about it. That does several things. First, it allows us to attract the best open-source talent. People want to come to Red Hat because they know everything we do is open. We release all of our code. Period. There’s nothing proprietary. I think that is meaningful to a lot of open-source developers so they want to be at Red Hat.
Second, it gives us a very crisp clear message in the market. We’re not saying open-source is good here, but then we go and do something else. We believe it’s the best way to develop software.
Then we give it away. It’s free. It’s up to us to figure out how to add value beyond the functionality of the software because it’s free. And since it’s fully open source, there is no vendor lock in.
We could just say “we are open core and here are few things that we will make proprietary.” But that dilutes the model a little bit. And frankly it makes you lazy, as you don’t have to then focus too much on adding value.
People often criticized the open source business model by saying there is only one Red Hat. Is that criticism valid?
The only other public open source software company now is Hortonworks and they’re pursuing our model. They’re totally like us; they’re totally open source. They’ve actually gotten quite a bit of traction and they’re growing nicely so we feel like our model can work in other areas.
I’ll be the first to say the market cap of proprietary software companies will never be matched by open source companies because part of the model is the IP is free. The vast majority of the benefit that open source creates is lower prices and more flexibility for end customers. That said, I think we will see, over time, maybe a dozen or so open source software companies. I don’t see why we wouldn’t see that over time. It’s just a matter of they need to mature and figure out their own business models.
Sam Ramji of Cloud Foundry once told me that in 5-10 years people won’t even use the term open source. It will be the defacto software development model. What’s your perspective on that?
I would say nothing against Sam, but I would be very careful about proprietary companies saying, “Over time we won’t even talk about open source.” There’s still a real battle between open source and proprietary. Most proprietary companies recognize that they need to adopt components of open source because it’s a better development model certainly for a whole set of software out there. I think we’ll be talking about open source for a long time or things won’t be open source.
I’m thrilled to see so many proprietary companies adopting components of open source, and I do think Sam is right in the sense that developers like open source, have an affinity for open source, and contribute to open source. So that’s the reason Microsoft is open sourcing .NET. I think we’ll see more and more and more of that. I would be really hesitant to say people won’t be talking about open source.
In your keynote at Red Hat Summit this week, you mentioned that we’re witnessing a 4th industrial revolution going on with machine learning and AI. Where is Red Hat in that revolution?
One of the biggest difficulties of being the CEO of Red Hat is that we are a company with more opportunity than resources. If open source was a proprietary technology and we had been the large share player of Linux as we are with RHEL, we would be generating 20 billion dollars of revenue instead of two billion in total for the company.
We would have multi-billion dollar R&D budgets to be able to go after big data and analytics and SDN and all these other areas, but we’re not. We’re Red Hat and we’re all open source. The price points are lower. While our production system is applicable across so many areas we can only do so much at once.
RHEL is certainly a very profitable product, along with many others. And we are investing heavily in OpenStack and containers, all of the components around that — from Docker to Kubernetes, the management stack, storage… It’s about all we can handle right now.
As some of those products start to mature a little bit we’ll certainly look to invest in other areas. What limits us are our financial and managerial resources to be able to go into those areas. There’s no reason Red Hat couldn’t be the big player in big data or the big player in SDN or the big player in analytics. It’s just a matter of where we are investing heavily right now. Those are the areas we are in; I can’t do five or six other things.
Red Hat is often being accused of too RHEL-centric. How would you respond to that?
We don’t think of ourselves as particularly RHEL-centric. I do think we are around OpenStack in particular. We do believe that there’s such tight linkage between the hypervisor and the operating system that it’s hard to break that thing apart, to engineer it in a way that we think we can offer production grade support on.
If you look broadly, our storage platform has nothing to do with Linux. We just acquired API management company 3scale and fully open sourced it. It’s not about Linux. What we’re doing around mobile is not about Linux.
Much of what we’re doing isn’t directly related to Linux itself and so we don’t necessarily see it that way. Now, I do think as things start to fragment out we see a distributed world where we could play a big relevant role in that.
There are some open source companies that don’t enjoy the same reputation within the community that Red Hat does. How do you manage to strike that fine balance between enterprise and the community?
We put community first. We have an upstream first policy. Everything we do goes upstream and that’s just a part of who we are and what we do. We always say there’s not a conflict between community and upstream because the communities are in enterprise, because we’re all in.
It’s in our DNA. We recognize it’s an ecosystem and so we understand the importance of giving back. I will give you an example of our very small work station business. We’re not deeply in the desktop business. Yet, we have the largest team doing virtually every major upstream GUI interface set of projects. Not because that’s so directly important to our commercial business but because it’s important to the Linux community. If it’s important to the Linux community, it’s important to Red Hat. So we invest in areas that aren’t necessarily commercial.
We don’t commercialize products that we’re not deeply involved in. When I buy an open source company, let’s say Inktank, and pay some hundred million dollars, I am often asked by the board of directors that it must have great intellectual property. And my answer is that it does: it’s Ceph, but it’s free and anyone can use it. Then they say that it must have a great brand and I say, yes, it’s called Ceph but anyone can freely use it and we are committing that it remains free. Then the board comes back and says it must have a lot of revenue and my answer is no, it has virtually no revenue. Then the board asks why are you buying it for hundreds of millions of dollars? My answer is “influence in the community.” We want to show that we have the key developers and that we are involved and committed to the community. There are others out there who offer open source products based on Ceph but do zero, squat for the community.
Another good reason for upstream first is that when an enterprise says that they need some features in there or if there is an issue that needs to be fixed, if there is a hotfix you should have that confidence that you can put it in upstream. If you are not contributing to that project, how can you really do that? So being a contributor is really important.
We believe that if we are really good contributors. If we really focus hard on giving back, not only the direct features that we need, but more broadly to the ecosystem, that’s going to be overall good for us.
We don’t hold things back. When we build things we don’t hide it from people to do a big reveal and say “here is the code.” We don’t do that. We work in communities from day one, all the time. I do think we try to be really good participants, and I think that comes through.
By doing that, I think we’re, hopefully, perceived as being good citizens in the communities in which we participate.