Migration to Containers, Microservices and Kubernetes

Mirantis CEO Adrian Ionel reflects on the state of Kubernetes two years after acquiring Docker Enterprise. Video is below followed by a transcript of the conversation with Mike Vizard.

Male:                          This is Digital Anarchist.

Mike:                           Hey, guys. Thanks for the throw. We’re here with Adrian Ionel, CEO of Mirantis. It’s been two years since the big deal with Docker that brought across all the technologies for the enterprise from those guys and put it in the Mirantis portfolio, and it’s been expanding ever since. Adrian, welcome to the show.

Adrian Ionel:              Oh, hey, Mike. Great to be here.

Mike:                           Where are we now as far as the migration to containers, and microservices, and Kubernetes, and all these platforms? I feel like we’ve been talking about it off and on now for five years, and every year we say, “This is the year.” But I wonder if it just takes longer than people think, or we’re just moving faster than people realize?

Adrian Ionel:              Actually it’s skyrocketing, and I’ll give an example. We have this product called Lens, which is very popular, maybe the most popular ID for Kubernetes, which Kubernetes is already easy to use. And to give you some data points, Lens had about 10,000 users 20 months ago. Today, it has 450,000 users worldwide. And they’re managing about a quarter of a million Kubernetes, individual Kubernetes clusters in all kinds of sized companies from world’s largest companies like Apple, or SAP, or AWS, all the way to a small indie shop somewhere in India or in Scandinavia. So that gives an idea for the adoption of Kubernetes at massive, massive scale. So yeah, it’s skyrocketing. It’s totally there.

Mike:                           Early on, a lot of this was driven by developers, and now I think there’s more IT operations people involved. So is this a bottom-up phenomenon still or is it being driven more by operations folks that like the efficiency that Kubernetes provides?

Adrian Ionel:              You’re absolutely right. Early on, it’s the developers who just moved to containers and said, “This is the way we would like to work.” But now we’re seeing more and more where even CIOs and VP of IT operations are just providing mandates into their infrastructure to embrace Kubernetes and containers as the standard because one is that customers are requesting it, which are the developers and the application owners, and second because it’s just so much more efficient. It leads to an infrastructure that’s much easier to operate. It’s standardized. It works across clouds and on-prem. So it has all kinds of benefits. So it’s both. It’s the developer _____, and it’s now also IT operations and IT leadership just mandating it as a new way of doing things.

Mike:                           As we go along here, do you think people are appreciating the complexity of some of these applications? If I look at microservices, there’s a lot of dependencies. If I look at Kubernetes itself, there’s a lot of knobs to turn and people use the phrase “wall of YAML piles”, and they get a little intimidated. So I know we promised people that things would get easier and that we’d all have flying cars by now, but are things getting more complex?

Adrian Ionel:              Yeah, you’re exactly right. There’s a lot of power in Kubernetes, and containers, and in microservices, but with that power comes a lot of complexity as in cloud services or cloud platforms in general. And in fact, this is one of the key problems that we’re solving for our customers. How do we take a technology that’s super complex, but very powerful, and turn in something that’s very, very easy for people to use, and very, very easy to operate.

So yeah, it is complex. But through technology, through automation, through software, the complexity can be removed or hidden from most people so that people can take advantage of containers and Kubernetes very, very easily, and not have the headaches of knowing every single detail.

Mike:                           What is your sense of how much of the infrastructure is going to be managed by an internal IT team, or a thing shifting to this kind of as a service or managed service approach because people want to spend more time writing code?

Adrian Ionel:              Yeah. Well, we are really biased there. We have our unique point of view. The philosophy that we’re taking as we approach every single company on the planet to become a zero ops, a zero ops business because we fundamentally believe that for the vast majority of businesses there, operations is not a place where you can create a ton of unique value. And even then maybe the best possible operations architects and engineers can be more focused on designing and crafting new services, new platforms rather than trying just to keep the trends moving and up and running.

So yeah, we believe that everything should be delivered as a service, including infrastructure that even on prem, people will want a true cloud experience and as-a-service experience with zero operations. And this is central to the value that we provide our customers. Not everybody is there, but we see more and more companies actually totally embracing this and shifting their most critical talent towards creating new services, creating new products, creating software versus running ops.

Mike:                           Early on, Mirantis of course was closely associated with OpenStack. What’s the relationship between OpenStack and Kubernetes these days? Sometimes you look at it, and Kubernetes can run on OpenStack. OpenStack can run on Kubernetes. The two run side-by-side. What, are people using both or is an either/or?

Adrian Ionel:              Yeah, the relationship could not be better actually. And this is part of our playbook in _____ for acquiring Dock Enterprise. Even before we acquired Dock Enterprise, we containerized OpenStack, and we ported it to run directly on Kubernetes just like any other microservices applications. So you can think of OpenStack really as a very complex microservices architecture and application. So right now the way we ship OpenStack to our customers in our latest release on our next generation product has been running on top of Kubernetes.

So Kubernetes is the substrate on which everything runs on our platform including OpenStack. And that gives customers a tremendous amount of flexibility and choice because in our own unified fabric, if they need a virtualization, they can put OpenStack on top of it. And then some customers, you’re absolutely right, put yet another layer of user centric or workload-centric Kubernetes on top. So you can have the sandwich where you can have Kubernetes as the base layer, OpenStack on top, and yet again more Kubernetes clusters on top of OpenStack. And that’s a very typical architecture that people run and give people a tremendous amount of flexibility in that.

Mike:                           We live in this multi-cloud world. Some of it is on purpose. Most of it is by accident. But we talk a lot about hybrid cloud, and Kubernetes seems to have a lot of potential in that space. Do you think we will see Kubernetes on-premise and in the cloud, and that will become the foundation for our future hybrid cloud ambitions?

Adrian Ionel:              Yeah. We definitely see Kubernetes on-prem and in the cloud, and a combination of the both. I’m not too sure that hybrid cloud as – you know, it depends very much what we mean by hybrid cloud. So we see very few cases where customers split on application between on-prem and public cloud. What we see much more frequently is that customers categorize their applications in cohorts. And certain cohorts have been moved onto public clouds. Other cohorts are being kept on-prem. But there is one common and unifying team which is that the vast majority of customers wanted the development teams to use open standards and cloud native standards for the development frameworks and for the infrastructure they use.

And this is where Kubernetes comes in extremely handy because that means that people that develop on applications for on-prem target really use the same frameworks and the same tools and methodologies that people who develop in the cloud. And it also means that the customers can move those apps wherever they want at any point in time with the relative ease and have a lot of portability. Which is why we believe that anything and everything that will remain on-prem as an infrastructure will eventually also migrate towards cloud native standards centered around Kubernetes and containers. So that developers and application owners have the same experience in a public cloud and on-prem.

Mike:                           Post-COVID, we saw a lot of spending on IT, and a lot of it driven by digital business transformation initiatives to varying degrees of success. Do you think we’re going to enter a new period though where there’s gonna be a lot more focus on the cost of IT and trying to contain that a little bit more because we did go through a period where almost out of desperation, we were spending on IT to hopefully drive reinvention, but maybe some of the old disciplines will be coming back.

Adrian Ionel:              I don’t know it’s a very, very good question. I feel that the emphasis is going to be on measuring the return on investment in a much more rigorous way rather than just be a cost. I think that companies will very much continue to invest very heavily on those digital initiatives that really deliver good results or how the promise of phenomenal results and are transformative for businesses.

And at the same time, they will question every single dollar that’s just spent on keeping the lights on, or any kind of dollar that is really not producing meaning for competitive advantage for business. So that’s going to – I think people are going to take a very hard look at that. And we know from our experience that a very large portion of IT budgets today are still being spent on maintaining legacy, maintaining legacy applications, maintaining legacy enterprise datacenters.

And that is going to be, I think, a target for cost savings and finding – freeing up capital to invest in new software development and new initiatives that can really transform the business. So I’m not sure who’s going to be a cost reduction as much as how can we actually ensure that we get a real return on the money that we spend on information technology.

Mike:                           We also hear a lot about securing the software supply chain these days. And I wonder if people have yet to fully appreciate this, but one of the benefits of microservices and containers is it’s easy to rip and replace something that has a vulnerability in it than it is to patch a monolithic application. So do you think we’ll get to a point soon where people will start building more in microservices because of security concerns?

Adrian Ionel:              That’s a very good question. We do have a lot of very security-sensitive customers throughout the federal government, Department of Defense, National Labs, that’s not just in the West, but in other friendly, US-friendly nations around the world. You have a lot of security-conscious customers as well, whether it’s banks, pharma companies, healthcare companies, and the like, payments services like Visa, MasterCard, PayPal support.

So we do believe that security is very much top of mind for anybody and everybody that’s running digital initiatives these days. And we do see absolutely that people truly realize that the container gives you just, I think the most elegant, the most safe way to patch vulnerabilities today versus any other technology that’s out there. And you’re absolutely right, that may yet be another driver for people wanting to containerize that they can respond much more quickly, and predictably, and reliably to security threats.

Mike:                           Right now, there’s no shortage of Kubernetes platforms and distributions, so ultimately in your mind, what differentiates one provider from another? What should people be thinking about?

Adrian Ionel:              Well, I do think there is a shortage or there is a very limited number of platforms that are actually proven at scale, and we fit into that category. If you have the 5G network at AT&T, that runs on our Kubernetes plus an open platform on top, and _____. Or you could take what we do from Visa, or many other large customers around the world, actually doing it at massive scale is challenging, and doing it securely, being able to update and patch, and operate, and guarantee results is quite difficult at the enterprise scale. And when you put that lens on, I think the number of players that can actually do it with a high degree of quality is very small. It’s less than you can count on one hand.

Mike:                             Do you think it will ultimately get easier? One of the criticisms you hear about Kubernetes in general is you’re confronted with this wall of YAML files and nobody knows how to cope with that. So ultimately, I wonder if people are still intimidated by the platform and maybe we need this whole level of abstraction above it, and that’s where we’re moving to.

Adrian Ionel:              That’s absolutely right. And that’s in part what we’re adding on and what you’ll see coming out from here on is more and more a focus on application owners and developers, and how can we move essentially all the complexity of Kubernetes or having to know something about Kubernetes from the developers. So the newest offerings that we launched, like DevOps powered by Linux, and that’s enterprise, are all about making it super easy for the application on developers to actually build apps, inspect apps, run apps, update the apps on Kubernetes platform without having to be Kubernetes experts.

And my personal view, and I know people have different views in the industry on this, but my personal view is that the average developer does not want to know anything about Kubernetes at all. They just want to build the code that meets their end users’ needs, and push it down to the platform. And everything else has to be done through automation.

Mike:                           All right. Well, this is the second anniversary. Assuming there’s a third, what are we gonna be talking about in a year from now? What’s your predictions for 2022?

Adrian Ionel:              You mean for us as Mirantis as a company?

Mike:                           Yeah, absolutely.

Adrian Ionel:              Yeah, well, I’ll tell you what we’re focused on. One of the things that – there are many things that happened that we’re super proud of, and I think the most important that happened _____ is just how we delivered for our customers, and for our people, for our colleagues from _____ company that that we’ve gone through the past few years as a successful and robust business continuing to grow and continuing to invest. And doing it entirely out of our own funding, the big cash for _____. So it’s been absolutely fantastic.

But one of the things that all of us as a team are most proud of is that we have a group customers, and a sizable group of customers, it’s not just five or ten, a very significant group of customers who have been growing at 400 percent per year on average with us over the past three years. So these are, we call them fast growers or hyper growers. And I believe what we will be talking about in a year from now is these hyper growers growing even faster, growing maybe not just at 4x a year, but maybe growing as high as 10x a year. That’s our ambition at getting them to a much larger scale, much, much faster, and having a lot more of them.

Mike:                           Very good. Hey, Adrian, thanks for spending some time with us.

Adrian Ionel:              My pleasure. All the best. Take care.

Mike:                           All right. Back to you guys in the studio.

Adrian Ionel:              Okay. Awesome. This was terrific. Thank you, Mike, for doing this.

[End of audio]

 

Mike Vizard

Mike Vizard is a seasoned IT journalist with over 25 years of experience. He also contributed to IT Business Edge, Channel Insider, Baseline and a variety of other IT titles. Previously, Vizard was the editorial director for Ziff-Davis Enterprise as well as Editor-in-Chief for CRN and InfoWorld.

Mike Vizard has 1662 posts and counting. See all posts by Mike Vizard