 In the beginning, there were mainframes, a highly centralized, secure, command and control environment. Open systems brought a spate of innovations that were powered by machines, servers, storage arrays, networks that had to be configured, deployed, and managed by specialists. No virtualization had made that simpler, but it was still a machine-centric world. The cloud, DevOps, and importantly, containers created an inflection point in the industry where no longer did developers have to do a handoff to an infrastructure guru to deploy and often reconfigure systems, which could cause other problems. Containers essentially codified the infrastructure to the point where developers could now be responsible for the full stack with consistency that allows stretching, if you will, of applications between on-prem, to the cloud, across clouds, and out to the edge. Kubernetes in particular has enabled organizations to host applications and containers with automation. So you can now deploy as many instances of your application as required and communicate between different services used by those applications in a consistent manner. What this does is enables rolling updates, security patches, and a run anywhere environment that is changing how organizations build and manage their applications. Hello and welcome to this CUBE conversation and preview to KubeCon CloudNativeCon North America 2021. I'm pleased to welcome my friend and guest, Stu Miniman, Director of Market Insights for Cloud Platforms at Red Hat. Stu, man, great to see you. So good to see you, Dave. Thanks for having me. You're very welcome. So you heard my little spiel up front, my little narrative. What are the big trends that you're seeing that you're watching that you think people should know about that are important? Yeah, well Dave, I'm so glad you started out talking about the application because Dave, I mean, you know my background, your background very much too, is started in infrastructure. And for so long we talked about, well, it's different increments that we talk about the infrastructure, but there was that huge divide between the people that run the infrastructure and the people that build and own the applications. And when Agile and DevOps came out, we talked about not throwing things over the wall. But when we look at containers in Kubernetes, really what it is is an application to build our application, to modernize our application, to run our application. As you said, they have to be more that right once go anywhere has been something we've wanted for the while. And from a developer viewpoint, I haven't wanted to think about the infrastructure. So we want to enable that. We want developers to be able to do their thing. What we've done at Red Hat is try to have that consistency in every environment because Kubernetes is only a very thin layer. There's lots that needs to be done on top of that. But one of the biggest trends is from an application standpoint. The same thing that we've seen in other environments Dave, when you say, okay, well what apps did you have? Well, it's great to say I have the cool microservice, new stuff, but what about older applications? What about modernizing things? Can I lift things over? Can I have a broader spectrum of applications? And yes, that's where we are with Kubernetes. We don't just have stateless applications that are written in this new modern way. We have a broad spectrum. And there's another word that I really keyed off of in your intro, talking about automation Dave. If you talk about scale and you talk about automation, that's what container was built for. If you look at what the predecessor to Kubernetes was Borg at Google. And if you think about just building things at scale and building things with automation at their core, that's what we've done. And that's where this ecosystem is building towards. So not saying everybody needs to be Google, but when you start talking about AI applications, when you start talking about different ways to really have automation built into your environment, this is where containers and Kubernetes really shines because that's where we've really gone beyond human scale Dave and we've gone to that machine scale. So we need to make sure not just to remove humans, to remove errors, but to be able to have that agility and flexibility and scale, which is what offers in this space. So all the cool kids, of course they want to develop in the cloud, but I feel like for every app that's developed in the cloud, there's like 10 on-prem that are screaming to be modernized. And we have a chart on this, but so what kind of applications are you seeing going in to containers and Kubernetes? Yeah, so two charts here. So it's a survey we actually did for QCon Europe leading up to it. The one on the left talks about the data. Is it stateless applications? Is it stateful applications? Well, what do you know Dave? It's a mix of both of those. You'll remember Dave, in the virtualization days, it took us about a decade to solve those storage and networking things. How do we make sure that things really run at the virtual machine layer? How do we have things like moving all over the place and still not break the connection that we had there? That was a lot of hard work that we as an industry did. Well, here we are six, seven years into Kubernetes. We've solved a lot of those same issues. So storage and networking work much better today in a Kubernetes environments than it did in the early days. It started out, oh, stateless applications, but if you look at the data on the second side, what kind of applications are there? The answer Dave is yes. You want your cool new modern databases, absolutely, AI and ML, absolutely, through kind of your ISV, more traditional applications. The answer is yes. So customers are doing a whole lot of it. When I'm meeting with customers, one of the first questions we always have. Dave, we've worked on silo busting for many decades in this industry, but if you talk to the infrastructure team and you ask them, well, what apps are you putting on there? If they don't have a good answer, the first thing we do is, hey, you really need to get the developers in the room. You really need to understand this because if you stand up a platform, just because Kubernetes is cool and it's great and it helps you build your resume, you're not going to have success down the road. You want to make sure they're involved upfront, understand what the requirements. So Kubernetes, one of the joke is, containers and Kubernetes adds a magic and yippy you win, it's like, well, there's a little bit more to that to actually have it work. You mentioned it took a decade plus to actually kind of work it out in the virtualization days that maybe you remember the API, you know, stuff. We have the scars from the remnants. Right, exactly, but it's interesting when I look at this chart that, you know, because like you said, it started off, it's kind of stateless database. Yes, all kinds of applications, but database is number one. And so you've got a lot of stateful applications, enterprise apps, security sensitive. I mean, everything's security sensitive today, but hyper security sensitive. So do you feel like that timeframe relative to, you know, two decades ago is going to be compressed? It seems like it's compressing quite rapidly. Absolutely, the CNCF always puts out a survey around these ventures to where adoption is. It's a little bit of a self-selecting for the community, but containers and Kubernetes brought adoption. We've really, not only have we crossed the chasm, we're into the, you know, solid majority of adoption here. And yeah, the databases, I mean, Dave, you've covered things like the Postgres world, companies like Crunchy Data, and some of these modern databases are really built for this type of environment. And as you said, they shouldn't have to think as much about, okay, I'm in a cloud or I'm in a different cloud, this containerized platform for applications can live in a lot of different places. And that goes to kind of what we're seeing changing in the infrastructure world over the last couple of years. Glad to mention that database, I was interviewing Josh at the Postgres event, and he was explaining to me how far Kubernetes has actually come and how much, you know, more trustworthy it is today, still some gaps, but much different than even two or three years ago. Yeah, I guess one of the highlights interesting at the KubeCon Europe, there was the general availability of both the Pipelines project and the GitOps project. It was, it's Argo CD is the project for GitOps. And when that went GA for Red Hat, we actually have that built into OpenShift at GA. And not only was it ready to go, we actually had a few customers that were ready to say, hey, we're using this and we're using the production. So we had XA Insurance, one of the largest payers in the globe, and the largest bank in Turkey, were two of the ones that we had saying, hey, we're using this. For the audience, if you're not familiar with GitOps, it's everything we use GitHub as the repository of records. So that this is kind of, if you think about the old days, we had the gold CD or the gold server. Well, we do that for our entire stack, that whole infrastructure's code that we've been talking about so many years, but it will manage that for us. So I patch it at the GitHub level and it will enforce what I have in my environment. So if somebody, oh wait, let me make a change. No, it's constantly validating things at GitHub. So it keeps it rather regimented. So we've had, as I mentioned, a couple of customers we've seen a lot of interest in the public sector space because of course, Dave, they're very concerned around security and patching and access. And we want to keep that least access necessary. So if we can keep that at the GitHub level, that's one of the things that will help your environment. It really ties into the whole kind of GitOps, AIOps, modern environment. So it really ties all of it together as to kind of the culture, the application and the infrastructure. So your files, your config files, your policies, same API, same console, that is how you get the scale. Yeah, absolutely. We don't want the people to have to manage that as much. So you can let them focus on where they're going to add value to the business. So let's talk about cloud. The definition of cloud is changing. Cloud is expanding, it's going on prem. There's hybrid connections to a cloud or multiple clouds, across clouds now seems to be becoming more real. We could talk about that and then maybe eventually out to the edge. They're all real in their own right, but how much is actually being connected together is something that I'm interested in. But what are you seeing there? What role is Kubernetes playing? Yeah, so first you talked about where applications live. The latest data I've seen from kind of the industry watchers is what are we, Dave? 20, 25% of applications are in the cloud. That means there's a lot still in the data center. If I look at OpenShift customers, yes, do we have a lot of them in the data center, but then they are also using the public cloud. So we have deep partnerships with Amazon and Azure to do public services in the cloud. And our value is we give consistency across all of those environments. So are you using data center? Yes, most customers still have data center. Do you have one or more clouds? Absolutely, I used to love the Andy Jassy line. Multi-cloud doesn't mean that you spread evenly across all the clouds. Most customers I talk to, they have a primary provider that they partner with, but things change over time. We've seen plenty of customers go two or three years in and say, well, I have a strategic initiative. Sometimes they make an acquisition and they will do another cloud. Or there's lots of factors why I might be doing more than one cloud. There's certain industries where basically you have to have relationships with multiple vendors or there's regulations that you need to be concerned about. So the answer is yes. What we've been talking about more than a decade at Red Hat is open hybrid cloud. And what does that mean today? You might have not have planned it out, but you're hybrid today and what are you going to be in the next decade? You're gonna be even more hybrid. So Edge, if we talk about it, everyone is talking about, one of the biggest trends here is how does Kubernetes go out to the edge even more? That consistency message that I talked about, where does OpenShift live? OpenShift lives anywhere that Red Hat Enterprise Linux lives. So RHEL, am I gonna have Linux out at these small environments without a lot of resources? What else are you gonna have other than Linux? That's going to be the foundation of what you have. So if I can have management and consistency that push out to all of those environments, and we've been building out a portfolio, something that you'll see us talking about more at KubeCon in LA is single node OpenShift. So this is a really small footprint OpenShift, but still have the consistency to work across all these environments. And we've had different footprints basically to be able to do Edge and remote offices, whether you're talking from a service provider out to a full customer premise data center, but there's a lot going on in the edge space. We actually have, we already have a public use case with Verizon who's doing some of the AI use cases. I'm sure you can picture with Verizon being such a large telco, the touch points that they have, not only at the service provider, but to their customer environments and OpenShift as the platform for enabling that innovation. I mean, if I had a big application portfolio on-prem, and the legacy company with 100 year history, obviously I'm going to be doing some stuff on the cloud, I would be building some kind of abstraction layer that would, could obviously modernize my on-prem estate. I would want to, I would probably start with Amazon. I'd want to take advantage of AWS cloud native tooling, but I would absolutely be doing the same thing in Azure and Google. And I would want to build my own cloud, right? And service my customers or my company, have people log into that cloud, hide the underlying complexity of the technology and just simplify everything, up level it and build a stack around that. I'd probably build it on OpenShift. Why not? And of course Kubernetes, but there are alternatives. There's EKS anywhere, for example, which presumably is a competitor. How is that impacting the marketplace? Yeah, so Dave, as you said, everybody is kind of extending beyond where they live. So Microsoft Azure has their Arc offering. Google has Anthos and Amazon was the last one. I mean, Dave, you'll remember this. When we talked about hybrid and multi-cloud for a bunch of years, it was like Amazon doesn't talk about hybrid or multi-cloud. And back when I sat on the analyst side, I was like, well, you can't talk about hybrid and multi-cloud without talking about Amazon. So they've now, EKS anywhere, something they announced back at Reinvent, it just went generally available recently. And so they have a distribution of Kubernetes that you can use on your own. So you could have completely disconnected in your data center, running only on VMware is the only way that they support it today. And they have in beta, there's something called an EKS connector. So if you want it to be managed from the cloud and have someone more of that consistency, they have the way to do that. They've had EKS, which is their Kubernetes service in Amazon for a bunch of years. But as a friend of the program, Corey Quinn says, there's actually 17 different ways to run containers in Amazon today that's supported by Amazon. And you laugh at it, but, you know, David, it's no different. You know, remember the storage world? Okay, how many different storage products did EMC have? Do you know how many compute and storage products Amazon have? They have a lot. It's growing. So one of those offerings that they have, natively in the console is Red Hat OpenShift Service for AWS. So is EKSD a competitor? Well, if you're an Amazon customer and you want everything Amazon and you want to use their environment in a hybrid environment, yes, you can do that. Part of the strategy for Amazon is Outposts. We've got on our roadmap to be able to support OpenShift on Outposts. So, you know, we look at our positioning is we are much more than Kubernetes. If you talk about the stack of tooling that we build on top of it, we've done a real lot to make sure that developers have the tooling that they need from an Amazon environment. It's just the Kubernetes piece. It's a, in the cloud, it's a managed control plane in your own data center. It's here's a Kubernetes distribution. Good luck with it. If you want monitoring and observability, if you want more security, if you want all these other pieces, you need to build them on top of that as opposed to OpenShift gives you a full application development platform. You know, Forrester Wave, we were, you know, far and away, the top and to the right on that spectrum with the leading position for both developers and operators. So, you know, great to see Amazon. You know, I hate to say they're like validating something that we do, but look, everybody's going to do this distribution everywhere. I know that's the marketing line, but it's... And I hate to do the marketing line, but it's, you see everyone rolls out their pieces and you say, what is the game that they are playing? It's Amazon, wants you to consume as much of their services as you can. From a Red Hat standpoint, it's, well, everywhere that RHEL can go, we can go. So, OpenShift can live a lot of places. We are going to give you the best experience in your data center, in Amazon, in Azure, in Google, in your hosted, in the Edge. We're going to work in all of those environments and we've got years of experience with thousands of production and employment, like in the data center, EKS anywhere sitting on top of vSphere. As far as I know, we have at Red Hat the most production Kubernetes deployments on VMware or OpenShift. Actually, at VMworld, I'll be talking about, I'm on a panel talking about OpenShift on vSphere with VMware. So, long, deep partnership that we've had there. No one can speak to the breadth and depth of what we've done there. What's the little lie? Amazon always says there's no compression algorithm for experience. Well, I like it. Okay, but that's why I like your Edge strategy because I've said many times the Edge is going to be won by developers. It's not going to be won by taking a, you know, x86 box, throwing it over the fence and saying, okay, we got Edge. And I think, you know, that's tongue in cheek. I think that the traditional enterprise hardware vendors are understanding that, but they're not in a great position with developers. You know, maybe Cisco a little bit with DevNet, but generally speaking, you know, VMware obviously always has been struggling. The Edge is, you know, the challenge with the Edge is you always have to look through it as to what your perspective is. So we have a long and deep relationship with a lot of the telecommunication providers. People will disparage OpenStack some, but that's actually the solutions that we've sold the most into our network function virtualization for the telco. And a lot of them have followed what they worked with us on OpenStack and continue that into OpenShift and Verizon being one of those proof points. You've seen my ETR data and I tell you OpenStack keeps popping up and when you dig into it, it's, oh, that's telco. There may not be, or maybe there's not a region there and it's telcos developing their own cloud, essentially. And you know, they're monetizing it. So let's talk about a CNCF, the ecosystem. We have another slide on this if you guys wouldn't mind bringing it up. I mean, it's a complicated matter, right? You got, here's the picture. I mean, you can't read it because there's just so many people there. What's the stop this from becoming, you know, kind of OpenStack too? Yeah, that's a great question. So Chris write our CTO, I thought really boiled it down really well. One of the big problems with OpenStack is we were building a complete stack. So when they said, oh, there's all these projects, it's like, okay, well, we're gonna create a big tent. And under that big tent, you have to have all of these pieces and they all need to work together. And while there were modular projects, I needed to have that full stack validated and managing and maintaining that was a nightmare. What is the CNCF landscape? It is, you know, dozens, 100 more projects that are independent of what they had. So yes, Kubernetes is the one that gets the most attention, but take something like ServiceMesh. ServiceMesh has been around for a few years, it's hot. We're still early on the adoption trend. ServiceMesh works with Kubernetes, but it isn't limited to Kubernetes. It's one of those Venn diagram. It works with it, but you can also work with my virtual environment, it works in other places. And that's true of a lot of these projects. Often they are complimentary to Kubernetes, but I can adopt them standalone. So the challenge is it is that paradox of choice when you go out there. There are some people that wanna go to the grocery store and buy all of their various pieces and put it all together. Well, other people will come to us and say, hey, I just want my developers to get working. I don't want them to spend all their time fighting over what they had. And at Red Hat, we say, great, we're going to have an opinionated platform. And if you come down later and say, oh, there's a piece of it I don't wanna use or I have some other tool I can have, it's betters are included, they're optional and they're swappable. So that's what's nice in this developer environment. So we also work with companies like HashiCorp. A lot of our customers use Vault for their secrets. GitLab is another peer of our industry that have a lot of developer tools. They're not a Kubernetes provider. They usually sit higher up in the stack than we do. So there's a lot of players, there's a lot of room for activity and innovation. Yes, we've seen a Cambrian explosion of projects there and there has been some consolidation that's part of the job of the CNCF is in the observability world. They took, I can't remember, there were two projects that were kind of similar and they got them in a room and got them to agree to put them into a single project and put those together. So we do see some consolidation over time, but there's still room for a lot of growth. Standards are good, but so is optionality, I think is your point there. So the event is October 11th to the 15th. It's actually an in-person event you're planning on being there. So I am, it's hybrid. I know a lot of people will be online. The other thing I'd point out there are a lot of day zero events. So these are really awesome. There's a GitOps day, there's security day, there's so many different pieces. I'll actually be for the day zero. I'll be emceeing the OpenShift Commons where we get a bunch of end users to just tell their stories, projects they're working on, deployments that they have, have some good partner ecosystem discussion there. It's usually a lot of fun. We hope a bunch of people come to those in-perses and then the day itself, the three days of the show itself are always hopping in, lots of learning to be done, whether you're there in-person or online. Fantastic, so I'm glad you pointed out it is a hybrid event. That's kind of the nature of these things these days and I think will be for some time. I think potentially and definitely, I think people are realizing, hey, you know what, as much of a pain in the neck as virtual events are, we can reach a lot more people and it's a good on-demand experience. Have at it. Stu, thanks so much for coming into the Cube Studios. We miss you, glad to see you're thriving and good luck at the show and we'll see you around the block. Thanks Dave, I know I'll be seeing John on the Cube there too. Absolutely, okay, thanks for watching everybody. This is Dave Vellante, we'll see you next time.