 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 hosts, John Furrier and Jeff Rick. Hey, welcome back everyone. We are here live in Silicon Valley where theCUBE lives in Palo Alto. This is, we're in Mountain View at the Computer History Museum. It's theCUBE, our flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier with my co-host, Jeff Rick here at the OpenStack Silicon Valley, hashtag OSSV15. Join the conversation. Lutaker, CTO, Cloud Computing, and Cisco. Legend in the industry. We saw them at OpenStack Summit. Good to see you again. Good to see you guys. Great to be back. Cube alum in the OpenStack. So, okay, cloud innovation. Silicon Valley's innovation. Intel's here, Cisco's here. All the big names are here. Big sea chain, big transformation. We've heard that. What's to meet on the bone? Tell us what's happening. What are you talking about? You know, of course, cloud computing, the internet is just continuing to move through different industry after industry. So one of the things I'll be talking about here at the summit is that how we've seen, you know, Airbnb is disrupting hospitality, you know, that we've seen Netflix disrupting, you know, traditional video delivery. Now we're seeing cloud computing going into other areas, such as in terms of network services with large telcos, mobile packet core. A lot of these things are changing the whole dynamics where we've delivered hardware, you know, for specific systems, video encoding, transcoding, video production has always been, for example, hardware-based that's moving to software. And OpenStack is coming in just at that transition. So OpenStack is going to be the platform all of those innovations start to happen. So I got to ask you, because it says on the website here, on featuring your thing, it says, quote, changing the face of service delivery. That's the tagline, it's the marketing slogan. Okay, you're the CTO. It's not a marketing slogan by any means. What does that mean? What does that mean? It's a very accurate statement about what's going on. It's actual marketing, good, we love the data. What does it mean, changing service delivery? So service delivery in many cases, like I mentioned, has been done by these hardened appliances. If you've seen large, large racks of networking equipment, or if you go into any TV studios, you see large, large racks of transcoders and coders, all of that big switching things, all of that's being disrupted now. So we're changing how the face, how that is going to be delivered to the customer. We've seen Netflix do that with DVRs basically in the cloud. Now we're going to start seeing a hit broadcast video. We're seeing it hit mobile packet core. So cell phone traffic, we'll start running on top VMs on top of OpenStack. So is this a function of performance? Is it a function of stability or both? That's a great question. It's really both, and it's really Moore's Law. Moore's Law is constantly moving forward here. And in fact, I actually have a, we're at the Computer History Museum. I have a machine I had part of the design I've called the Connection Machine. It's one of the dinosaurs downstairs. That machine, which was the world's fastest machine for about a year, has been replaced. And Moore's Law continue, and now it's actually can be replaced just about by your laptop. It's kind of like when good enough is good enough, right? Because this building used to be, before it was the Computer History Museum, it was an SGI building, I believe. That's true. That's true. Maybe a headquarters at some point, or that was the big one over across the street when they went bananas with that headquarters. So again, specialty boxes, specialty microprocessors, specialty OS and software, which eventually, good enough, got good enough to displace that. And actually it's even better than good enough. Because I think that when we see these things move into software, you can upgrade these things instantly. Instead of having a truck roll to come out and replace a piece of equipment, you can download a new virtual machine or the containerization of these software services so that you can immediately get the, so you get an acceleration effect from the fact that you move from hardware-based systems to software-based systems. So I got to ask you, you brought up the Computers Museum where one of your original designers was the fastest thing around for a while, got replaced by the next big thing. That's Moore's Law, it was totally awesome. That's what clouds bring to the table. But here at the History Museum, there's a Google self-driving car in the lobby, which I love to look into because they're driving around, I can't tell, they're taking photos, I'd say peace or I was just someone in there. But that really illustrates this new software paradigm because Internet of Things is about the self-driving car. It's about real time. So how does that tie in? Because that's a great example of the kind of innovation we're seeing. The Internet of Things was the wearable watch where I get my tweets and crowdsets to my wrist to self-driving cars need self-healing. They need real-time data. They need real reliability, not, you know. And they need this both in the car to do the near term navigation, but then they're also getting all the information from the cloud. So they're getting a larger view. And this is where we're seeing, you know, the history of computing has always been actually this between decentralized, distributed computing, to centralized things. Now I think what we're seeing is the growth on both sides. So that we're seeing services being delivered from the cloud in the case of IoT, managing perhaps, you know, sensors and collectors and everything else that are on-premise. So we're seeing the union of those two things so that we get the best of cloud computing. So I think that in terms of OpenStack, we're going to see an evolution now of OpenStack, whereby OpenStack will start to reach out beyond the data center. And it'll reach out so that you're having management capabilities, orchestration services. So service delivery is also the full end-to-end from that user's device across the internet, across multiple data centers, to wherever they're trying to get. So this service is now going to be a generic size, generic in the definition. It could mean any service. So for instance, the self-driving car will rely on heavy compute on the device, the car, through a cloud delivery of some services, reliability because you got to be real, real-time. You can't be like a half a second. You're driving a car. No, you have to be responsive. And that's why just like the human body works that way, we've got sensors and flexors and reactions that are local, even though they're mediated then by a higher function that are in the cloud or in our heads. The other thing that I love about the self-driving car is they published a lot of data recently on statistics and information that these things have collected over whatever, 1.3 million miles. And it's the second order kind of impact where just in the way that they're collecting data to move through the environment, they're now taking a snapshot of that data around them. And they did a whole thing about bicycles because they're seeing crashes and bicycle behavior that they're not necessarily involved in, but they still can rebuild the scenarios across multiple instances to come up with new rules in ways to operate. You know, I think the biggest challenge will be us as humans accepting the automation that that brings because there's a lot of information showing that it'll be much, much safer than having humans drive cars. But then we feel very nervous about handing that over to a computer system. So fully automated. Is it going to have a steering wheel or not? What if, you know, like we have with airplanes today which are largely driven on autopilot most of the time and that a lot of times the problems arise when you actually give humans the ability to jump in and do something, they often do something wrong. So we have to find the balance how humans can stay in the loop, but we still hand more and more over to the automaker. I love talking to you because you've, again, you've been an inventor, you've been a developer, you've done some things that are even in the history museum here. So it's great to kind of get your mind around around this. I got to ask you the software question. You know, we hear orchestration of containers. There's so much more work to be done. You think about just concepts of geospatial data, how drones going to be delivered to unique locations, to real-time information on the car, internet of things, all this powered by the cloud. Where are we in this? So, how early are we? I mean, is it at just anything connected to a network? So one of the things that I think that we have to get used to almost is we've talked about continuous integration where we're taking new components, putting in continuous delivery, where we're actually deploying multiple times a day of these new services. We used to talk about continuous learning. This is a challenge for our developer community. There is so much that is happening now and the developers are, I think, having to keep up with now containerization. Containers are great because they make it easier for the application developer. So we will be making more and more tools are continuing to come out. And we're seeing more and more tools than being applied in cloud computing. So these services, we're talking about microservices, there's other ways to construct application. So we need to become continuously learning so that we can keep up with all of this. And the devices too, but I want to ask that question on the learning. That kind of implies that if it's moving so fast, and this is a trend that I'm seeing, I've been fleshed out publicly at the first time, is the developer communities themselves are optimizing how they talk to each other and you see the movement of how fast Docker has risen. That kind of teases out this notion that there's a lot of reliance on each other now as developers more than ever. That's great. It's a great point because I think that we're particularly, we continue to move into the way that applications are being constructed as they are assembling different pieces of different open source pieces of software and services provided by multiple providers. It's an assembly process. This is great. This is what we do in manufacturing plants. You know, the automobile manufacturer doesn't build every piece of that component. They farm that out and they assemble these things. So that's what we're seeing with software now as well, I think. And you got to trust your partner. It's almost like being in a cohesive military operation or some sort of team approach because you got to know your partner who's got your back is going to be trusted on his code or multiple sources of that. You need, there's a lot of things that you need but it is becoming an assembly. So that will allow us, I think, to build more reliability because we can actually have the different components compete for the business of being in that manufacturing pipeline and that the developer's job becomes one of assembly and of architecture. A lot of people look at you as a mentor certainly with your history and track record in the business. Is there an issue that you're pounding the table or, you know, hey kids, get off my lawn. I mean, is there a thing that you're so passionate about that you want to share with the folks out there that you can just pass on just a tidbit of like. I think we're almost in that golden age of development now for application developers. It's a great time to be a developer. The number of tools that we're talking about are available. The amount of software that's available in open source. You just have a plethora of things that you can use. So I think it's important now for the developers to start taking the user's point of view with their trying to develop something for somebody else to use and they can draw upon this enormous tool set. So it's a great time to be a developer but it does require continuous learning. You've got to be up on your skills. You've got to take advantage of all of these innovations and not be locked into a single way of looking at it. They have to be able to look at things in a much more cohesive, you know. What about the bubble discussion? We were talking to a venture capital earlier and I would say you need to comment on whether you think it's bubbles, maybe you will or not. But for the young generation that haven't lived through multiple cycles of innovation, what's your advice to them? I mean, do you give a caution? You throw caution to the wind, you say go for broke, break glass, spit out nails, what are your advice? The bubbles are applied to financials. I don't see bubbles applying to technology. What we're seeing in technology is a relentless, you know, continuation, like I said, of Moore's law of tools and everything else. So if developers keep true to what their skills really are about, and that is innovation, technology and everything else like that, they can ride through any bubble that makes them. Lou, what about the buyers? I always think of the poor buyers. Because they're getting hit with this plethora of options. Everything's getting cheaper. It's getting cheaper, but it's changing so quickly. You know, I just kind of got a dupe underneath. I kind of got a dupe figured out enough, and then I got to do Spark. Right, right. I mean, just the rate of change, and then with so many options on the open source side, and the way those things are growing, from the buyer perspective, he's just trying to, you know, run his business. It is a lot more difficult, I think, on the buying side. And for the CIOs, a large organization, having to make these things, I think they have to make bets. You can only succeed if you're willing to make a bet and you stick to it. And that's why I think OpenStack right now is getting to be a safer and safer bet. Open source gives us multiple vendors, so that by with OpenStack, you're not dependent upon a single vendor anymore. We have independence of that. You're not being locked in. So the buyers get this freedom when they're going with a platform such as OpenStack. But they got to make that bet. Great comment about innovation about riding out the bubble. Certainly, I totally agree with you. Technologies underpinnings are being re-architected, so just ride it out, don't get distracted. Next thing, networking. Talk about technology and innovation. Real quickly, as we end the segment, you're going to go on panel. Certainly, the services thing we hit on already. What's going on in networking? Where are we? Update, still under construction? Is there any progress? It's interesting, the last couple of years we've seen two big, I think, sort of things that are coming together as perfect storm. We've seen cloud computing, and we've seen software-defined networking. The movement, as we talked earlier, about two software functions instead of specialized hardware appliances. Those things are converging now, and we're seeing that called network function virtualization. I think it's really just the virtualization of all these services, whether they be thought of as a network service, or a media service, or a transcoding. This is where it's becoming, because at the end of the day, it's always about networking. We are moving information through our networks from a source of content development, or a video stream, out to a user. We're moving in between data centers. So networking is continuing to evolve, I think as we have these new demands that are coming in. Okay, so I got to ask you a question. I know we're getting the hook here, but I want to get one more question in. Craig Mclucky from Google, we were talking earlier, he made a point that I liked, which was in talking about API economy. If you forced the API application of the enterprise too early, you could really taint some of the value that could be created from some of the legacy stuff. So the same question comes to you around NFV. If we force NFV or virtualization or SDN too early, is there a foreclosure situation or tainted goods? Well, people will get burned. And where is the hotspot? And well, I think that you want to, there's a couple of things, Jonathan Bryce actually made a couple of important points at first in kicking it off, which was that you have to have a focus and you pick where there's going to be the highest value in a very simple area that you can succeed in. Build on success. If you take on more than you can consume, more than you can deliver, you're going to be risking failure. So focus in on the high value teams that are moving fast that have a very simple objective and then go behind that success. All right, Luke Tucker, thanks for your time. I know you got a panel. Okay, good luck with the panel. Great to get you on the front end of that live. I was looking at Luke Tucker. He's got something here in the museum and winning and he was displaced by Moore's law, but he's here with Cisco doing more. Great to see him on theCUBE. Thanks for sharing the data and wisdom. We'll be right back more live in Silicon Valley here at theCUBE after this short break.