 in the computer museum in the heart of Silicon Valley, extracting the signal from the noise. It's theCUBE covering OpenStack Silicon Valley 2015, brought to you by Morantis. Now your host, John Furrier. Okay, welcome back everyone. We're live in Silicon Valley for theCUBE, our flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier, the founder of Silicon Angle. We're here live at the OpenStack Silicon Valley event. It's really hashtag OpenStack SV or hashtag OSSV15, join the conversation. Our two next guests are Tim LaBelle, cloud product manager at HP, Hewlett Packard and Scott Miller, senior technologist at DreamWorks feature animation. Welcome to theCUBE. Thank you. So I love when DreamWorks is always involved in theCUBE and I'll see a big HP customer. But you guys have been, I've seen on stage multiple times at an OpenStack event, also other industry events. You guys are always pushing state of the art, whether it's kick ass workstations and any kind of tech to give you guys competitive edge. You guys are always doing something cool. Yeah, that's exactly what people think we're a content provider. We do feature animation, we do television, we do consumer products, but behind all of that is some amazing technology. A lot of stuff exciting. So here the cloud is a new thing and with cloud you could do a lot more, a lot more compute, you got storage, more flexibility. So with all that promise, you sure you got some budget, you're kicking the tires, you're doing some deployments. Share with the audience what are the things that you guys are doing with OpenStack and how are you guys rolling out some of this cutting edge stuff? We've actually been doing the cloud for almost 10 years depending on your definition of cloud. Our first production KVM virtual machine went live in 2007. We were using Zen before that as our hypervisor. Been using HP Public Cloud for storage for a worldwide collaboration for about four years now. The thing I'm most excited about I think is OpenStack as a control plane is going to give me a single common way to manage all my storage, my networking, my compute. It's vendor agnostic, it's self-service and it's automated. Tell me about this control plane because that's the notion that if you think about it, it makes a lot of sense, but most people get confused that OpenStack's a platform, OpenStack's a distribution. Can I buy OpenStack? Can I buy distributed computing? I mean, so they're all concepts versus reality. So what is the reality of this control plane, OpenStack and how do you talk to your peers about that? From our perspective, OpenStack is a collection of APIs, really good ideas, architectures and implementations. So by control plan, I'm referring to the parts of the system that let me schedule compute machines, schedule some storage, allocate a network route, orchestrate the creation and removal of object storage, any of those things. Most shops have a control plane like our legacy control plane. It includes the telephone and the ticketing system as part of the APIs. Someone opens a ticket, says, I'd like a new VM, please. Some number of hours later, a few days later, someone's created a VM, provided it back to them and they're done. With the OpenStack control plane, I remove all that human middleware and I can go direct to the end user if they need a virtual machine. They can run a Nova command and create a virtual machine. Okay, so let's talk about the reality of OpenStack. So to the keynote today and HP, you guys can comment on this. It was really refreshing because we've been bullish on OpenStack for a while, but this part's OpenStack that's emerging that's not ready for prime time. Certainly this stuff, and Jonathan laid that out today. Mature technologies that are winners, virtualization, we know that check, but OpenCore, I mean Core Compute for OpenStack, the compute piece is solid. And what else is coming up that's moving fast into this winner category of OpenStack? Yeah, so I work a lot with customers at HP working with the OpenStack components and you're right, they've now found a solid foundation that they're building on and they're starting to throw some small workloads on there, but they really want to push it even further. And for me, the area of containers and the area of focusing on the application and not as much on the networking and the infrastructure, but that requires you organizationally to be ready to do that as well. And so what we find is companies are challenged with, it's not just a technology challenge, it can often be an organizational challenge. Tim, you've got Scott here as a customer. You're in the product management side, which is really tough with OpenStack. I can almost imagine because every use case is different. I mean, beauty is in the eye and the beholder dream works as probably different use cases than other customers. How do you sort through the product management nightmare and how do you talk to the suppliers out there and say, hey, I don't really care. I just want good cost to buy, good capex, op-ex, and I want mobility workload management. I got to do my job. So whatever you want to call it, hybrid cloud, this, that, or the other thing. How do you sort that out and how do you talk to the suppliers? For sorting it out, we create a partnership with a set of our customers and I'm involved deeply in that program at HP. And we choose partners such that we go across a lot of different sectors, a lot of different workloads, a lot of different use cases to make sure that we've got coverage as much as possible. And then it's really a partnership. I mean, we sit down and we have arguments. They put down their requirements. They're pretty clear about what they want. So we find companies that are not only technologically advanced, but they're also leading edge and dream works fits that. How do you deal with that? You know, if you're a product management in school, in business school, it's like the standard thing. We don't do one-offs. We want broad market requirements for all of our customers. That was the old way. Now, in a way, we're living in a world of one-offs. How do you talk to the suppliers? We're not actually looking for one-offs. I'm looking for one-off ways to use the standard thing. The more eyeballs and the more people in the ecosystem, the better OpenStack's going to become. I mentioned earlier that OpenStack's a collection of ideas and APIs and architecture. There is an implementation. I think probably the biggest challenge right now is the implementation on how do you install it? How do you maintain it? How do you keep it up to date? And how do you operate a nonstop enterprise on top of an ever-changing ecosystem? That's probably been my biggest collection of feedback. As I mentioned, we've been an HP partner for more than 15 years now. And we've always gotten to them because they not only like to solve problems, they have to solve hard problems. I think HP funds nine or 11 of the 16 or so PTLs. So they have access to the engineering staff that we need to fix problems right away. They've got a great professional services organization that understands installing and operating OpenStack. I mean, one of the things we're struggling with, it's more of an opportunity than a problem is, how do I get my developers? It's the cloud problem in general, but it's the OpenStack cloud problem in particular. How do you design applications? How do you develop them? How do you deploy them? How do you debug them? And HP's helping us try to figure that out. It's a lifestyle change for us. I had to tweak that out because it was an epic comment. I love that comment. We want to do one-offs of standard things. And that ultimately is what OpenStack's all about in my mind, because we hear that all the time for people in the industry is that it's a collection of Lego blocks, whatever metaphor you want to use, but it comes down to engineering. This is the DevOps ethos that's now coming into the enterprise. So Lisa Marie's book, OpenStack Technology, breaking the barrier in the enterprise, little plug for Lisa Marie, there's your plug, little note in there for me. The enterprise adoption is about the standardization. That's the anti-lock in. If you believe that to be true, which we do, what is the engineering challenge that practitioners have to take? Because it's not just developing software, and software is fundamental, but software development is one element of engineering. You got the operations side and step off. So how do you sort that out? How can you share your opinion on this whole engineering challenge that enterprises have? Yeah, that's one bit of the challenge is we've got a lot of legacy software development, skillset organized around long running projects with a waterfall type schedule, and convincing and demonstrating the benefits of continuous deployment, continuous integration, and the DevOps mentality, it's been a challenge. A lot of software developers don't want to know how to log onto the machine, they don't want to understand how to provision virtual hardware, they don't want to wire up their own network, and to the degree we can, we hide that from them, but we're encouraging people to develop in the same exact runtime environment where we're deploying in production. So if it breaks, or when it breaks, they have a place to debug, they'll understand the runtime. So you were doing cloud before cloud was cloud. Yeah, you should be called utility computing. Grid. It's been called grid computing, it's been called off-site computing, it's had a lot of different things. Distributed computing. All of those things. Computing? Yeah, and cloud is a marketing, architecture turns how to rules all of that together. It is distributed computing. I think the fundamental difference between what used to be called cloud and open stack is that I can dynamically provision my infrastructure, whereas before I might have dynamically deployed my application in a relatively static infrastructure. So the ability to grow and shrink machines and storage and networking, and then those other components and reaction to my workload, that's huge. Okay, so I've got to ask you the question. Then you get both of the answers. So hybrid cloud, is there a category of hybrid cloud? Do people buy hybrid cloud, or is that kind of like distributed computing where it's a mindset? Because certainly private cloud, you have on-premise dynamics. Costs, legacy, applications, skills. You got public cloud for resources, but the workloads really are agnostic about no one goes to hybrid land. I mean, is there a hybrid product, or is just the totality of public-private working together the notion of a mindset? So is there a hybrid cloud category? I don't think it's a product. I think it's a way of using the existing products. And we've got on-premise open stack. We've got managed private cloud open stack, and we've got public cloud open stack. So that's a hybrid, and any one application may have a foot in all of those areas, or may only be in one of those areas. And the goal for me and our whole organization is the agility that lets me pick. If I've got a latency sensitive, rapid deployment type application, I might do that on-premise. If I've got something with maybe relaxed security requirements and needs of global reach, I'm going to put that in public cloud. If I've got something where I want to have better control over the infrastructure and tendency, I'm going to put that in managed private cloud. But the company's hybrid, but any one app might only be in one place. The totality is a hybrid environment, but the core fundamental stuff is going on. Tim, how do you address that? So let's talk about that, because now you've mentioned containers, right? You've got containers, you've got Kubernetes, you've got these new tooling, huge growth and push and tooling, just floating on the scene. Right, I think hybrid is a thing. I think definitions are all over the place. I think Scotty's got it right on. There are multiple, there's managed private clouds, there's private clouds, there's public clouds. You need to look at your application and pick the right environment to put it out on. But standards are critical, like the open standards that we find with OpenStack and so forth are critical, because you don't want to be changing your application, but you still want to have the flexibility to go between clouds. To me, that's hybrid, and that's what people are asking for. I'm going to ask you guys both the question. You mentioned architecture, vaporware. These are terms that we kick around. Marketing, buzzwords, cloud, big day. We know what they all are, but it really speaks to the high level underlying trends. But I got to ask you guys the more fundamental question is, abstraction layers are critical in all this, right? So developers are key, how you roll this out. Hiding the complexity has always been a fun engineering challenge and challenge for the vendors who are doing the best to win, right? So talk about what's going on there to hide the complexities, to make the deployment and the management, the development easier and simpler and more elegant. I would say one thing is that as you move up the stack and you get to the developer layer, developers want to have access to things such as being able to provision their infrastructure and so forth. Once you start getting there, that is complicated to some of them. So what companies are doing is they're trying to find ways to either standardize within their company how do we develop software, or they'll create another abstraction layer to help their developers fit into that. But the goal is that from managing the cloud standpoint, you need to maintain those standards. Yeah, so what? Abstraction layers, what do you guys do it from tooling? Sure, you're going to do a couple things. The tooling and the continuous integration, continuous deployment part, it's still for us, it's very much a work in progress. But one of the things we are trying to do is, there's always the question of how do you take your legacy applications and put them in the cloud? So we're taking pieces of our legacy applications and instead of deploying them statically on pre-provision machines and having the application aware of how to talk to itself, we're deploying each piece, we're trying to microservice some of the legacy applications. So we're deploying the components in containers and each application thinks it owns the machine, but it's in a container with some nice walls. And then we're wiring that together externally so that I can mix and match containers on machines and create my application. And each of the developers is contributing their piece without really being aware of the rest of the infrastructure. Scott, tell us about this event. Tell us about this event. What's going on here in Silicon Valley? From your perspective, you're a practitioner, you're a customer of HP, but more importantly, you're a technologist. You care about the future, certainly. You don't want to be poisoned and toxic with vendor lock-in and all that stuff. We want choice, open source is powering everything. How do you look at what's happening in OpenStack in Silicon Valley here and also beyond, what's the status? Is it mature? Are you happy about it? What's going on from your perspective? I'm certainly happy about the direction. An event like this is a clear case. This is a Wednesday and Thursday in the middle of the week and the place is packed. People care enough to come here to talk about OpenStack, talk about their deployment, their problems, meet the vendors. And then the other aspect of it, it's a very social, software development activity. You're live tweeting, I think Lisa's live tweeting. I'm live tweeting while we're on live. Crowd chatting. In the old days you'd wait six months or a year to get some vendor announcement. Here it's pretty much tweeted and this code's being written in the back room and they'll release it by five. It's a very dynamic, interactive environment between vendors, operators, and consumers. How has that changed your life? Because we were talking in the Cube in Seattle last week, our lot are on the ground Cube, which is with open source now, I mean I'm going to be 50 this year so I remember when OpenStack hits a scene so I love it, right? But I still think there's going to be a whole nother level of open source because of the communication's now frictionless. The workgroup's now virtual, standardized around the world. This is a new model. Do you see it go into a whole nother level? Yeah, I think the expectation is everything is going to be dynamically reprovisioned, re-implemented, redeployed at a really rapid pace. I mean people, you talk about DevOps and a lot of DevOps development organizations have a cadence of maybe two weeks or four weeks. I think you're going to find continuous evolving software that gets releases every day. In fact, there are software applications today to get updates every day. It's maybe dangerous, but it means you get a lot of eyeballs. The developers are the operators, so if they make a mistake, they fix it. It means you're less likely to make a mistake. And I think it's going to make for healthier, more dynamic software. The scary part is, as an operator or as someone trying to train users, it's almost impossible because a feature you had yesterday might have evolved or changed. So we have to get used to software that's not static. Real-time dynamics. Tim, I'll give you the final word getting the right amount of time here, but talk about HP with the split now happening kind of pre the November 1st date. Do you guys feel like relief? The cloud's been an emerging group within HP. Certainly your contribution, I hope it's been very solid, like number one on overall contribution of the numbers you talked to. You guys had a big customer base. You feel energized? You feel like the team's coming together? Give us some insight. There's a lot of momentum going on right now. And what's fun is working with the customers is they've moved from just trying this out and kicking the tires to starting to put workloads on it, which makes things break down a little bit, which requires us to go back and improve the product. But our investment in the community for OpenStack is strong. We'll continue to grow on that. And one of the areas I think that is going to change in open source is we've got enterprise customers who are building cool code and wanting to move that along. And we're going to help them get that into the open source community. Well, it's been fun to watch the cloud group and watch you guys go through. And you had a lot of, you know, obviously distraction with the company stuff, but good to see you guys kind of being focused. We're looking forward to seeing you guys, hopefully in London that HP discovers what you've got. Thanks for sharing your perspective. How are you? Welcome. DreamWorks, that's a great, great place to work. A lot of action. Final question, what's the coolest thing that's going on in DreamWorks right now from a tech perspective? Oh, from a tech perspective, I think we're trying to reinvent how we deliver things like image generation and asset management for our characters as dynamically provisioned cloud environments as opposed to our old school legacy environment. And of course those characters can be retailed out in any virtual environment. You got to track those assets now all over the place. Holograms to social. A lot of good stuff. Thanks so much. It's a cue. We'll be right back here in Silicon Valley live after this short break.