 from Seattle, Washington. It's theCUBE, covering KubeCon and CloudNativeCon North America 2018. Brought to you by Red Hat, the CloudNative Computing Foundation and its ecosystem partners. Hey, welcome back everyone. Live Kube coverage here at KubeCon, CloudNativeCon 2018 in Seattle. I'm Chef Horace Dude Miniman, hosting three days of coverage, wall-to-wall, 8,000 people doubled from last year in North America, expanding into China, Europe, everywhere. The CNCF is expanding, so is Kubernetes. The rise of Kubernetes has spawned the CloudNative movement going mainstream. That's ecosystem-driven. We got a great guest here, Stephen Bauer. Data and Analytics Infrastructure Lead at Bloomberg featured them on SiliconANGLE.com in one of our special reports. End user using Kubernetes and the variety of CloudNative. Stephen, welcome to theCUBE. Thank you for having me. Thanks for coming on. Award-winning end user. Given all the end users, everyone's kind of award-winning. Yeah, yeah, yeah. Congratulations. Bloomberg's known, we've covered you guys, great development team. You guys have a lot of engineers at Bloomberg as well as being a media company on cable, Bloomberg Terminal, everything else. You got a lot of data science, you got a lot of engineers, you're building stuff. What's the focus on Kubernetes? Where are you using it? How are you contributing? What's the dynamic? Why are you winning with Kubernetes? Sure, that's a good question. I think, well, we're using it all over the place and lots of different things, right? We have a huge engineering team that does all kinds of different things. So in the area that I manage, data analytics infrastructure, we've basically managed databases and search engines and all kinds of other tech like that. What we've ended up realizing is that we built something that looks a lot like Kubernetes, but doesn't work nearly as well for all of those different systems to manage them at scale, right? So we're talking thousands of instances of Postgres and Solar and all kinds of different things and having a single tool or a single platform which we can kind of level up all of those things really makes a lot of sense in terms of not necessarily cutting costs and things like that, because that's actually not as interesting to me as actually allowing the teams that manage those things to actually contribute to those projects, contribute to Solar or Postgres and stuff like that and free them from having to spend a lot of time managing infrastructure. Tim Hocken said it was just on the cube here before you came on from Google, one of the co-leads on Kubernetes at GKE, Google's at cloud. He said, something interesting, I want to get your reaction to this. One of the benefits of Kubernetes is to give the confidence that deployments are going to be reliable and that confidence gets a flywheel and then people start shipping more as a matter of course of the business, not like, oh my God, we got to push a new code. Oh my God, fingers crossed, trust the button. The old model was, you know, fingers crossed, go, QA, no, no, confidence. The confidence and the iteration. Is that where you see in the value too? Does that relate to you? Does that make sense to you? Does that resonate with you? Yeah, I mean, it definitely does. A lot of the models that we're trying to move to is a really declarative model of both how we develop software and then how we deploy software and then how we manage it in production and Kubernetes kind of offers that ecosystem across the board and that's been really, I don't think of a great way to put this, but being able to have that tool and being able to do that and the repeatability in the world that I live in, everything we do, we don't do one of it. We do, I think we run something like 2,000 solar clusters. So all we're doing all day long is just stamping out the same thing over and over again and if I can build one system that does that really cleanly and simply and then I can use that same system for running Postgres or running something else, that gives us the confidence. We can test it, we can run it on our laptops. Our developers can develop and do all that kind of stuff and it works the same everywhere they go and we can just rinse, lather, repeat. So Steve, I want to step back for a second. Your infrastructure, is this all Bloomberg data centers? Is it, how does cloud fit into the discussion? Yeah, I mean, we do have some infrastructure running in the cloud, but primarily it's all on-prem and data center. In my world, it's all on metal because we have all these data systems that need direct access to SSDs and NME and all this kind of stuff. Can you give us, without sharing safe secrets, a little bit of the scale of what you're doing? I love data's at the center of what you're doing there. We all understand how important data is to your business but talk about what the requirements are that, is to, why you have some special requirements that the typical enterprise wouldn't. Sure, I mean, I think, you can look at Bloomberg as a media company. We have news, we obviously have the Bloomberg terminal and really what drives that terminal, it's all kinds of software, but in the end it's data, right? And it's all kinds of data. And that, what is the definition of big data and all these whatever stuff that everyone was pitching five years ago. We have all of those problems. We have data that is moving at millions of ticks a second. We have enormous data sets. We have really complex data sets like, oh, people scanning court filings from tiny little courts all around the country and sending that data in and we have to normalize that and put it into, so all these crazy different types of information, so they are both demanding in terms of the complexities of parsing data and putting them and structuring them into those systems as well as the scale. So we have some pretty enormous and high performance systems that require us and kind of drive us to that need for the metal and very focused on performance in all different aspects. Great, I wonder, give us your engagement with this ecosystem here, you know? One of the big questions coming in is, okay, so Kubernetes, the thing we hear from the CNTF is, well, it's getting kind of boring and I don't know that I agree with the term. I understand they're saying it's becoming mature and therefore there's less drama around it, which is good, but this ecosystem is anything but boring and you ask a user like yourself, you've got complex requirements, there's like more than 30 different projects here. What do you use out of here? What do you build yourself? What do you contribute to? How do you consider open source contributions? So it's a big nut and we don't have a ton of time but if you can scratch the surface on some of those. I think the number one lesson that I've learned from this ecosystem is that it's moving so rapidly that when we decide to build something on our own, we have a talk tomorrow about our data science platform which we built about a year and a half, two years ago. By the time we were ready to talk about it and everything like that, you have all the different technologies that have moved forward. So it's made us realize that if we're going to start something internally, a new project, either A, we should go look and see what's out there and contribute to that, or we should just start it in open source to begin with, rather than like, oh, let's build it and then we'll open source it. Chasing your tail kind of thing. Yeah, it's like, we have to become part of the ecosystem like in our entirety. That's pretty much a good question. I want to ask you this in context of thinking about your peers that might not be as progressive as Bloomberg on the tech side. You guys certainly do a great job and well documented. Classic IT shop, I mean, racking and stacking servers and boxes and now we've got the whole digital transformation thing going on, same old, same old. But now, 2019, real impact. The investments they're making on how to change their IT, their data is now in front of them. They have to deal with it. This is the right front and center because companies are realizing they're going to go out of business if they don't actually make the adoption. The data is super valuable. So how do you see the Kubernetes and the CNCF ecosystem changing the investment practices of a classic enterprise IT? Your peers called you and said, hey, Stephen, hey, help me out. What's the secret playbook? Where do I go? I don't want to get you out. I got to make some changes. And what do they change? What's the impact of the investment with Kubernetes? What's the real impact? I think, I mean, it's a tough thing, right? Cause Bloomberg is really not like your typical IT shop, right? We are a software company at heart, right? And so that makes us a little bit different when I talked to other people. I say that in the sense that like, not a lot of companies can afford to like decide to make a product open. Because they outsourced everything. Right? They outsourced everything. I think that's actually a huge change though, right? Like we're not sitting here talking about hundreds of commercial products that are, you know, owned by a small handful of vendors that are multimillion dollar investments for everything we're doing. We're talking about lots of little, tiny companies that have products that are really, really valuable, that are in the open source world, that we can get our hands on and start working with before we even make a decision about talking about support or, you know, whatever. There's all kinds of technologies that, you know, I walk into this room and, you know, these are like, they're like friends all around. Cause like we've worked with all the software and we're like, hey, these guys have a company now. This was just a GitHub repo a couple of years ago. And I think that that's a big change, like embracing that. I think that's probably really hard for your typical kind of IT shop where they want to have this clear line of, you know, I can call tech support and get some out on the phone. And that's like the main, you know. The classic old software model. Yeah. But it's changed. Yeah, so Steve, one of the things we're trying to get some insight on here is it's not just running Kubernetes in production. It's what am I doing with it? How does that change my business? Understand ML is a big piece of what you're doing there. Give us some insight as to, you know, how does this transform your business? Does it transform your business? I mean, specifically on the ML side, and like they'll talk about this, actually that's kind of the focus of our talk tomorrow. So I don't want to steal their thunder too much, but a lot of it was really about looking at, you know, okay, how did, you know, ML, like deep ML people, you know, work, how did they want to work, right? You know, if you ask an ML person what they really want, they want like an infinitely scalable cluster that they, it's just theirs. And they want an essay to manage all the infrastructure for them and a data engineer to manage the, you know, cleaning up all the data and all these things. And they wanted that all to themselves and not have to share it with anyone else. And so, you know, a lot of what we tried to figure out is like how we can actually deliver that to them. And it really is transformed. Like, you know, once people realized that on our platform, they had access to an enormous pool of GPUs, it went from being like, oh, I want to work on my box and can you give me GPUs on my one little box to, wow, I can, you know, do hyperparameter tuning across, you know, hundreds of GPUs, you know, overnight or, you know, during the day or whatever, you know, their needs are, like, it really unlocked people's like capabilities. And they're actually like, they went from kind of being skeptical of a system that they had to share and things like that. Because it actually just works. Like that, and that's really the- That's really the dopamine effect for them. Where you can see value without having to go through the slogging of the configurations and the normal stuff that they had to do. Yeah, authentication, you know. So we've been hearing threads of, you know, the CICD pipeline is a big benefit which you're kind of seeing as well. Below, we're also seeing people building below, Kubernetes, seeing storage and networking, getting better. How do you see that holistically? I mean, you're seeing as network more performant, the notion of programability becomes now part of it, automation, software, right? So now everyone has to build software. In fact, I talked to the VP of Technology Innovation at Proctor & Gamble and he's saying, hey, we outsourced everything, I got to start hiring software. So maybe not as big as Bloomberg, but the trend is let's get more software people on board. But they still got network, they still got storage, they still got the gear. What's the impact under the hood? Yeah, I think it's complex, right? Because you typically have these structures that are built inside companies, right? Where you have a networking team and you have infrastructure, hardware team and whatever. One of the SREs on my team the other day, he was like, do you think we can talk to the network team about putting software on their switches? Right, like that's a really interesting question to start asking. He actually had a really good use case, like that makes a lot of sense, like maybe we should think about that. And then dealing with, there's obviously the technology aspect of that, but there's also like skill sets, right? Like someone that's been working with a bunch of switches for a bunch of years isn't necessarily a programmer, right? It's used to a typical CICD process and things like that. And the other flip side, well, that's cool that he recognizes the networking guy, but we heard Tim Hopkins say there's a lot of policy knobs in Kubernetes that the networking guys could potentially take advantage of. So it might work the other way. Are the network guys looking at Kubernetes saying, hey, are they not yet that sophisticated? But, I mean, they would love, they love policy. Network guys love policy. I mean, come on, wouldn't you want to? It's actually one of the biggest draws of using Kubernetes in our ecosystem. We've made heavy use of applying network policy down to the workload level, which means that from a security perspective, if I know that I'm transiting data between two different places and I've only opened up assets for that one application, for that one particular use case, rather than saying, well, I know that I'm running the same workload on the same box and I got to open it up for everyone in that box, but maybe someone might use that thing, but maybe they won't and worrying about stuff like that. It's like, no, I can run a workload and I know that these are the only two endpoints that it can talk to. Oh, that's a relief. I mean, that's like, hey, we're done. For them, this is their panacea. I know exactly what workloads are doing exactly what on the network and what they're capable of. So that's been, especially for our data science workload. That's real progress. That's progress. Oh, it's huge progress, yeah. I've been able to do things that we used to not be able to do for years. Talk about the, okay. I just had a quick little question there. You mentioned you've got an SREs there. When did you pick that up as a term that you call there? And how do you see kind of, if you talk a little bit to the skill set and the jobs of people that you have inside? Sure, yeah. Boomers are a big company, so the terminology of it and what actually individual teams are doing is probably a little bit varied across the organization. It's been something that's come in over probably the last two to three years at Bloomberg. In my organization, it was actually really interesting because when I started off with, you read the Google book and whatever. What I did is I went to the guys on my team that were going to become the SREs for the organization and I had them write this manifesto about how we should build and deploy and manage software. And I didn't tell them necessarily up front that this is what was going to happen, but when they finished writing that and agreed that this is how things should work and they argued for a while, I said, okay, now go build all the tooling to make this easy for people to do, right? And that's what we, and then they've just been building out the tooling. Turns out like when you're working with a lot of the tools in the CNCF and then with Kubernetes, that's actually not that hard. Like there's lots of things there that are just easy when you get to that place. And so like that's the kind of journey we've been on to really try to build that infrastructure and they've done a good job and the engineers downstream of them have the speed that they're able to develop and the assurance that like, there was a CVE for Kubernetes two weeks ago and we patched it the afternoon that the CVE came out. Like being able to do that and any sort of company of scale is, I've worked a lot of banking and stuff like that in my past and it's unheard of to be able to deploy things in that speed. And that's really, I mean, this is the goodness of clouds, the goodness of having, you know, that kind of consistency operationally. It's funny you use SRE, it's a Google term, it's a great term. You got developers, you got operations, kind of working together now that's the magic. Well, Stephen, thanks so much for sharing this great insight on theCUBE, certainly great value for the folks watching. A lot of traction, a lot of people and users contributing and consuming Kubernetes, building around it, great trend. It's really fun to watch, a lot of composable services up and down the stack. So congratulations. Steve Bauer, Data Analytics Infrastructure Lead at Bloomberg, it's theCUBE bringing you all the action. Sharing the data here at KubeCon, this is theCUBE, we'll be right back with more after this short break.