>> Live from Barcelona, Spain, it's theCUBE! Covering KubeCon and CloudNativeCon Europe 2019. Brought to you by Red Hat, the Cloud Native Computing Foundation, and ecosystem partners. >> Welcome back. This is theCUBE's live coverage of KubeCon, CloudNativeCon 2019 here in Barcelona, Spain. 7,700 here in attendance, here about all the Cloud Native technologies. I'm Stu Miniman; my cohost to the two days of coverage is Corey Quinn. And to help us break down what's happening in this ecosystem, we've brought in Jason Bloomberg, who's the president at Intellyx. Jason, thanks so much for joining us. >> It's great to be here. >> All right. There's probably some things in the keynote I want to talk about, but I also want to get your general impression of the show and beyond the show, just the ecosystem here. Brian Liles came out this morning. He did not sing or rap for us this morning like he did yesterday. He did remind us that the dinners in Barcelona meant that people were a little late coming in here because, even once you've got through all of your rounds of tapas and everything like that, getting that final check might take a little while. They did eventually filter in, though. Always a fun city here in Barcelona. I found some interesting pieces. Always love some customer studies. Conde Nast talking about what they've done with their digital imprint. CERN, who we're going to have on this program. As a science lover, you want to geek out as to how they're finding the Higgs boson and how things like Kubernetes are helping them there. And digging into things like storage, which I worked at a storage company for 10 years. So, understanding that storage is hard. Well, yeah. When containers came out, I was like, "Oh, god, we just fixed it for virtualization, "and it took us a decade. "How are we going to do it this time?" And they actually quoted a crowd chat that we had in our community. Tim Hawken, of course one of the first Kubernetes guys, was in on that. And we're going to have Tim on this afternoon, too. So, just to set a little context there. Jason, what's your impressions of the show? Anything that has changed in your mind from when you came in here to today? Let's get into it from there. >> Well, this is my second KubeCon. The first one I went to was in Seattle in December. What's interesting from a big picture is really how quickly and broadly KubeCon has been adopted in the enterprise. It's still, in the broader scheme of things, relatively new, but it's really taking its place as the only container orchestrator anybody cares about. It sort of squashed the 20-or-so alternative container orchestrators that had a brief day in the sun. And furthermore, large enterprises are rapidly adopting it. It's remarkable how many of them have adopted it and how broadly, how large the deployment. The Conde Nast example was one. But there are quite a number. So we turned the corner, even though it's relatively immature technology. That's the interesting story as well, that there's still pieces missing. It's sort of like flying an airplane while you're still assembling it, which makes it that much more exciting. >> Yeah, one of the things that has excited me over the last 10 years in tech is how fast it takes me to go from ideation to production, has been shrinking. Big data was: "Let's take the thing that used to take five years "and get it down to 18 months." We all remember ERP deployments and how much money and people you need to throw at that. >> It still takes a lot of money and people. >> Right, because it's ERP. I was talking to one of the booths here, and they were doing an informal poll of, "How many of you are going to have Kubernetes "in production in the next six months?" Not testing it, but in production in the next six months, and it was more than half of the people were going to be ramping it up in that kind of environment. Anything architecturally? What's intriguing you? What's the area that you're digging down to? We know that we are not fully mature, and even though we're in production and huge growth, there's still plenty of work to do. >> An interesting thing about the audience here is it's primarily infrastructure engineers. And the show is aimed at the infrastructure engineers, so it's technical. It's focused on people who code for a living at the infrastructure level, not at the application level. So you have that overall context, and what you end up having, then, is a lot of discussions about the various components. "Here's how we do storage." "Here's how we do this, here's how we do that." And it's all these pieces that people now have to assemble, as opposed to thinking of it overall, from the broader context, which is where I like writing about, in terms of the bigger picture. So the bigger picture is really that Cloud Native, broadly speaking, is a new architectural paradigm. It's more than just an architectural trend. It's set of trends that really change the way we think about architecture. >> One interesting piece about Kubernetes, as well. One of the things we're seeing as we see Kubernetes start to expand out is, unlike serverless, it doesn't necessarily require the same level of, oh, just take everything you've done and spend 18 months rewriting it from scratch, and then it works in this new paradigm in a better way. It's much less of a painful conversion process. We saw in the keynote today that they took WebLogic, of all things, and dropped that into Kubernetes. If you can do it with something as challenging, in some respects, and as monolithic as WebLogic, then almost any other stack you're going to see winds up making some sense. >> Right, you mentioned serverless in contrast with Kubernetes, but actually, serverless is part of this Cloud Native paradigm as well. So it's broader than Kubernetes, although Kubernetes has established itself as the container orchestration platform of choice. But it's really an overall story about how we can leverage the best practices we've learned from cloud computing across the entire enterprise IT landscape, both in the cloud and on premises. And Kubernetes is driving this in large part, but it's bigger picture than the technology itself. That's what's so interesting, because it's so transformative, but people here are thinking about trees, not the forest. >> It's an interesting thing you say there, and I'm curious if you can help our community, Because they look at this, and they're like, "Kubernetes, Kubernetes, Kubernetes." Well, a bunch of the things sit on Kubernetes. As they've tried to say, it's a platform of platforms. It's not the piece. Many of the things can be with Kubernetes but don't have to be. So, the whole observability piece. We heard the merging of the OpenCensus, OpenTracing with OpenTelemetry. You don't have to have Kubernetes for that to be a piece of it. It can be serverless underneath it. It can be all these other pieces. Cloud Native architecture sits on top of it. So when you say Cloud Native architecture, what defines that? What are the pieces? How do I have to do it? Is it just, I have to have meditated properly and had a certain sense of being? What do we have to do to be Cloud Native? >> Well, an interesting way of looking at it is: What we have subtracted from the equation, so what is intentionally missing. Cloud Native is stateless, it is codeless, and it is trustless. Now, not to say that we don't have ways of dealing with state, and of course there's still plenty of code, and we still need trust. But those are architectural principals that really percolate through everything we do. So containers are inherently stateless; they're ephemeral. Kubernetes deals with ephemeral resources that come and go as needed. This is key part of how we achieve the scale we're looking for. So now we have to deal with state in a stateless environment, and we need to do that in a codeless way. By codeless, I mean declarative. Instead of saying, how are we going to do something? Let's write code for that, we're going to say, how are we going to do that? Let's write a configuration file, a YAML file, or some other declarative representation of what we want to do. And Kubernetes is driven this way. It's driven by configuration, which means that you don't need to fork it. You don't need to go in and monkey with the insides to do something with it. It's essentially configurable and extensible, as opposed to customizable. This is a new way of thinking about how to leverage open-source infrastructure software. In the past, it was open-source. Let's go in an monkey with the code, because that's one of the benefits of open-source. Nobody wants to do that now, because it's declaratively-driven, and it's configurable. >> Okay, I hear what you're saying, and I like what you're saying. But one of the things that people say here is everyone's a little bit different, and it is not one solution. There's lots of different paths, and that's what's causing a little bit of confusion as to which service mesh, or do I have a couple of pieces that overlap. And every deployment that I see of this is slightly different, so how do I have my cake and eat it, too? >> Well, you mentioned that Kubernetes is a platform of platforms, and there's little discussion of what we're actually doing with the Kubernetes here at the show. Occasionally, there's some talk about AI, and there's some talk about a few other things, but it's really up to the users of Kubernetes, who are now the development teams in the enterprises, to figure out what they want to do with it and, as such, figure out what capabilities they require. Depending upon what applications you're running and the business use cases, you may need certain things more than others. Because AI is very different from websites, it's very different from other things you might be running. So that's part of the benefit of a platform of platforms, is it's inherently configurable. You can pick and choose the capabilities you want without having to go into Kubernetes and fork it. We don't want 12 different Kubernetes that are incompatible with each other, but we're perfectly okay with different flavors that are all based on the same, fundamental, identical code base. >> We take a look at this entire conference, and it really comes across as, yes, it's KubeCon and CloudNativeCon. We look at the, I think, 36 projects that are now being managed by this. But if we look at the conversations of what's happening here, it's very clear that the focus of this show is Kubernetes and friends, where it tends to be taking the limelight of a lot of this. One of the challenges you start seeing as soon as you start moving up the stack, out through the rest of the stack, rather, and seeing what all of these Cloud Native technologies are is, increasingly, they're starting to be defined by what they aren't. I mean, you have the old saw of, serverless runs on servers, and other incredibly unhelpful sentiments. And we talk about what things aren't more so than we do what they are. And what about capabilities story? I don't have an answer for this. I think it's one of those areas where language is hard, and defining what these things are is incredibly difficult. But I see what you're saying. We absolutely are seeing a transformative moment. And one of the strangest things about it, to me at least, is the enthusiasm with which we're seeing large enterprises, that you don't generally think of as being particularly agile or fast-moving, are demonstrating otherwise. They're diving into this in fascinating ways. It's really been enlightening to have conversations for the last couple of days with companies that are embracing this new paradigm. >> Right. Well, in our perspective at Intellyx, we're focusing on digital transformation in the enterprise, which really means putting the customer first and having a customer-driven transformation of IT, as well as the organization itself. And it's hard to think in those terms, in customer-facing terms, when you're only talking about IT infrastructure. Be that as it may, it's still all customer-driven. And this is sometimes the missing piece, is how do we connect what we're doing on the infrastructure side with what customers require from these companies that are implementing it? Often, that missing piece centers on the workload. Because, from the infrastructure perspective, we have a notion of a workload, and we want workload portability. And portability is one of the key benefits of Kubernetes. It gives us a lot of flexibility in terms of scalability and deployment options, as well as resilience and other benefits. But the workload also represents the applications we're putting in front of our end users, whether they're employees or end customers. So that's they key piece that is like the keystone that ties the digital story, that is the customer-facing, technology-driven, technology-empowered story, with the IT infrastructure stories. How do we support the flexibility, scalability, resilience of the workloads that the business needs to meet its business goals? >> Yeah, I'm really glad you brought up that digital transformation piece, because I have two questions, and I want to make sure I'm allowing you to cover both of them. One is, the outcome we from people as well: "I need to be faster, and I need to be agile." But at the same point, which pieces should I, as an enterprise, really need to manage? Many of these pieces, shouldn't I just be able to consume it as a managed service? Because I don't need to worry about all of those pieces. The Google presentation this morning about storage was: You have two options. Path one is: we'll take care of all of that for you. Path two is: here's the level of turtles that you're going to go all the way down, and we all know how complicated storage is, and it's got to work. If I lose my state, if I lose my pieces there, I'm probably out of business or at least in really big trouble. The second piece on that, you talked about the application. And digital transformation. Speed's great and everything, but we've said at Wikibon that the thing that will differentiate the traditional companies and the digitally transformed is data will drive your business. You will have data, it will add value of business, and I don't feel that story has come out yet. Do you see that as the end result from this? And apologies for having two big, complex questions here for you. >> Well, data are core to the digital transformation story, and it's also an essential part of the Kubernetes story. Although, from the infrastructure perspective, we're really thinking more about compute than about data. But of course, everything boils down to the data. That is definitely always a key part of the story. And you're talking about the different options. You could run it yourself or run it as a managed service. This is a key part of the story as well, is that it's not about making a single choice. It's about having options, and this is part of the modern cloud storage. It's not just about, "Okay, we'll put everything in one public cloud." It's about having multiple public clouds, private clouds, on-premises virtualization, as well as legacy environments. This is what you call hybrid IT. Having an abstracted collection of environments that supports workload portability in order to meet the business needs for the infrastructure. And that workload portability, in the context of multiple clouds, that is becoming increasingly dependent on Kubernetes as an essential element of the infrastructure. So Kubernetes is not the be-all and end-all, but it's become an essentially necessary part of the infrastructure, to make this whole vision of hybrid IT and digital transformation work. >> For now. I mean, I maintain that, five years from now, no one is going to care about Kubernetes. And there's two ways that goes. Either it dries up, blows away, and something else replaces it, which I don't find likely, or, more likely, it slips beneath the surface of awareness for most people. >> I would agree, yeah. >> The same way that we're not sitting here, having an in-depth conversation about which distribution of Linux, or what Linux kernel or virtual memory manager we're working with. That stuff has all slipped under the surface, to the point where there are people who care tremendously about this, but you don't need to employ them at every company. And most companies don't even have to think about it. I think Kubernetes is heading that direction. >> Yeah, it looks like it. Obviously, things continue to evolve. Yeah, Linux is a good example. TCP/IP as well. I remember the network protocol wars of the early 90s, before the web came along, and it was, "Are we going to use Banyan VINES, "are we going to use NetWare?" Remember NetWare? "Or are we going to use TCP/IP or Token Ring?" Yeah! >> Thank you. >> We could use GDP, but I don't get it. >> Come on, KOBOL's coming back, we're going to bring back Token Ring, too. >> KOBOL never went away. Token Ring, though, it's long gone. >> I am disappointed in Corey, here, for not asking the question about portability. The concern we have, as you say: okay, I put Kubernetes in here because I want portability. Do I end up with least-common-denominator cloud? I'm making a decision that I'm not going to go deep on some of the pieces, because nice as the IPI lets things through, but we understand if I need to work across multiple environments, I'm usually making a trade-off there. What do you hear from customers? Are they aware that they're doing this? Is this a challenge for people, not getting the full benefit out of whichever primary or whichever clouds they are using? >> Well, portability is not just one thing. It's actually a set of capabilities, depending upon what you are trying to accomplish. So for instance, you may want to simply support backing up your workload, so you want to be able to move it from here to there, to back it up. Or you may want to leverage different public clouds, because different public clouds have different strengths. There may be some portability there. Or you may be doing cloud migration, where you're trying to move from on-premises to cloud, so it's kind of a one-time portability. So there could be a number of reasons why portability is important, and that could impact what it means to you, to move something from here to there. And why, how often you're going to do it, how important it is, whether it's a one-to-many kind of thing, or it's a one-to-one kind of thing. It really depends on what you're trying to accomplish. >> Jason, last thing real quick. What research do you see coming out of this? What follow-up? What should people be looking for from Intellyx in this space in the near future? >> Well, we continue to focus on hybrid IT, which include Kubernetes, as well as some of the interesting trends. One of the interesting stories is how Kubernetes is increasingly being deployed on the edge. And there's a very interesting story there with edge computing, because the telcos are, in large part, driving that, because of their 5G roll-outs. So we have this interesting confluence of disruptive trends. We have 5G, we have edge computing, we have Kubernetes, and it's also a key use case for OpenStack, as well. So it's like all of these interesting trends are converging to meet a new class of challenges. And AI is part of that story as well, because we want to run AI at the edge, as well. That's the sort of thing we do at Intellyx, is try to take multiple disruptive trends and show the big picture overall. And for my articles for SiliconANGLE, that's what I'm doing as well, so stay tuned for those. >> All right. Jason Bloomberg, thank you for helping us break down what we're doing in this environment. And as you said, actually, some people said OpenStack is dead. Look, it's alive and well in the Telco space and actually merging into a lot of these environments. Nothing ever dies in IT, and theCUBE always keeps rolling throughout all the shows. For Corey Quinn, I'm Stu Miniman. We have a full-packed day of interviews here, so be sure to stay with us. And thank you for watching theCUBE. (upbeat techno music)