 The Cube at OpenStack Summit at Lata 2014 is brought to you by Brocade. Say goodbye to the status quo and hello to Brocade. And Red Hat. Here are your hosts, John Furrier and Stu Miniman. Okay, welcome back everyone here. Live at OpenStack Summit in Atlanta, Georgia. I'm John Furrier, the founder of SiliconANG. I'm Joe Mike Coates, Stu Miniman. This is The Cube. It's our flagship program. We go out to the events to extract the signal from the noise. Our next guest is Lou Tucker, VP, CTO of Cloud Compute with Cisco Systems. Also vice chair of the foundation. Welcome to The Cube. Thank you. So you guys had a late board meeting last night into dinner. It was a late dinner. Big activity. I mean, the foundation now has done a good job of kind of putting the heat shield up a couple of years ago. You saw everyone jumping into one big marketing event. Oh, OpenStack's going to be this thing. And so you saw a lot of cloud washing. Peak cloud washing days a couple of years ago. And then OpenStack clearly went in the straight and narrow on just kicking ass, taking names, code, and use cases. We actually always hoped it would be like this. And we felt that by bringing together a lot of the experienced veterans in the industry who are on the foundation, representing their companies and everything else, like we could guide it in this way, with a very strong technical component. So it's always been that combination of those of us on the board, but the real strength of the community, it's in the code that is being written by the contributors, you know, day in and day out making OpenStack better. So you're talking to Eileen Evans. Just was on earlier with HP, deputy counsel with HP. She's on the board with you. You guys know each other from the Sunday. So I got to ask you the question. I mean, the cloud is one big distributed computer. If we're in an API economy, in a way it's a large supercomputer. It's a large mini. Well, you know, having built massively parallel supercomputers back in thinking machines days, and then many, many years at sun, we always knew the network is the computer. And so now actually, we're realizing that that is becoming a one big, the internet is one big computer. And certainly is happening right now. So like Scott McNeely said in the Cube, I should have just called it cloud. So I got to ask you, systems guys are back. I mean, you know, you've seen a lot of the, you said older, older guys, they should kind of come in together. That's kind of what's happening. Linux set the table as a, you know, as a chief alternative to the, to the proprietary stuff back in the day. Now Linux is the first class citizen in the enterprise. Now Linux build on that. And here we are. So systems is back in vogue. If you're a systems guy, it's good, right? Oh yeah, yeah, it's cool. Now, I mean infrastructure is cool. And actually people are asking, you know, a lot of the questions about it. And so that I think that we're seeing a resurgence in people looking at it from a systems point of view. And now it's at all the different layers. Because then that's what OpenStack is about. I think it's creating a truly unique layer in the new software stack for the data center. So just like middleware years ago, that we had, everybody said there's middleware. Here we have TIPCA, we've got app servers and everything else. Now I think it's this cloud platform, is that new middle, middleware that's moving into these things. And OpenStack is in the, is in the lead. So Lou, can you give us a little bit of insight into the inner workings of what's going on in OpenStack? You know, it's been, Icehouse was just released. So we're up to the alphabet up to I. That's right. We've made it easy for people by trying to keep it alphabetical. Do you know the next? Yeah, absolutely. But you know, people say, okay, we've been working on this for what, four years or so, and why isn't everything done? And you and I were talking offline that there's a bit of grumbling in the community that like Neutron, the networking stack, has some challenges and has some problems. So what's so difficult about what we're trying to do? We've taken on a really big challenge here. This is to build a in open source, a very new cloud platform that anybody can take and deploy their essentially their own Amazon in-house, you know, for a variety of different use cases. So the use cases span everything from a cloud service provider that we see, for example, HP and Rackspace and others providing those kinds of things, to somebody running it like Comcast who wants to run it on Xfinity. Those are very different sort of use cases. So that we've been fortunate because I think we structured the process around projects. And those projects are a compute project, NOVA, a storage one, SWIFT, a networking one around Neutron. That allowed the experts to get together. If it was all one big thing, we wouldn't have made anywhere near the progress that we'd seen. We still have a long ways to go. I mean, I think that we're seeing very rapid adoption, but it is by people who actually know they have to take on quite a bit. What we're trying to do, which was even mentioned this morning in the keynote, is a company like Comcast now is a contributor. So this is, I think, the new phenomena that companies are not looking to just go to a vendor and get a piece of software and install it and let the vendor do all the work. They actually want to be a part of this. And that's why we're seeing a lot of the service providers involved in an open source community. That's actually the first time I've seen service providers be actively involved in a community effort like this. So Lou, when I think of Cisco, I think about the people that have built their career off of being a Cisco administrator. You've got this army of CCIEs in the life. Exactly. It's a very different skill set than the folks that you see at this show. You're one of the leading people at Cisco that helped bring Cisco to the community. I think we want to take that whole community of CCIEs and we have to move their careers along. When they hear that they're actually, that they can, with a few lines of Python or Puppet scripts or whatever, start to orchestrate all the things that they were doing manually. They love it. It takes training. It takes education. And we are actively working with the CCIE community to educate them about this phenomena. And because it is open, it's freely available. The source code is there. You can read it. It's accessible to them. Lou, talk about what's different about this new philosophy because what you're essentially talking about is DevOps meets thinking machines, meets sun. You mentioned a member of the digital world. You were thinking machines. Yeah. Remember that startup in Cambridge, wasn't it? 65,000 processors all working in parallel. Just amazing. But back then, that was cutting edge. But now, these young developers, they come in, oh, I'm just going to get some Rails and some MongoDB and I'm good. I'm going to build a hyperscale. You might be able to maybe do some hacking, but to bring this system's mindset in there, what do these guys have to do? How do you take DevOps and cross them over to mainstream productive developers in this new framework? Well, DevOps is that intersection of software getting involved in these operational processes that we used to do by having a run book and you'd have a sequence of steps that some system administrator would go through in order to provision a new employee in their database. Now, we're recognizing that script. That can be made into a script. So we're seeing actually the between Puppet and Chef and Python. We're seeing kind of resurgence in these easier to use scripting languages, even though Python's going far beyond that, that are accessible in the people who used to do these things manually. With reuse too. With reuse. That's the whole point, because otherwise you can't, we want to take that, what one administrator used to do a thousand times and now he writes it once and he presses a button a thousand times. What's going to change in open source in your opinion and just look around the corner? Obviously, as open source matures and matures and matures, you know, some practices, certainly the ethos has never changed, but like in terms of mainstream, what do you see changing? In fact, I think OpenStack is a new kind of open source project in which the original code development is done in the open. This is not something that was developed as a product that was then decided to open source it. So to have a community of this size working on this bigger project with over a thousand contributors working together on these projects, just the way we manage that process itself is becoming something of an art. And so I know that you had Monty Taylor on here before he's running the infrastructure project, being able to say, how do we get all this stuff tested? How do we get all of these releases built? How do we integrate all of these projects? We are using automation. It's all about automation, automation, automation. We have to have the tools so it's all... You're writing the modern book of software development in this kind of way. And it may even change the way that we deploy applications. You don't get it on CD anymore. You'd be pointing to a GitHub repository. Versioning control, everything integrates. That's right. Lou, about a month or so ago Cisco launched InterCloud. I'm wondering, can you help us squint through that press release as to what it is and how does that fit into the whole OpenStack movement? Absolutely. We recognize, I think, by Cisco, first of all, we have a lot of SaaS applications. And many of the companies that we will probably continue to bring into the Cisco pool will be SaaS-based. So first and foremost, we started as an attempt to say we need a place to run all these SaaS applications. And many of those, for example, WebEx was already being ported to run on top of OpenStack. And so we started to make OpenStack the foundational element for this. We then decided to go further because to have one... Is Cisco actually hosting those SaaS applications? At a Cisco facility? Yes, absolutely will be. That'll be a part of InterCloud. So part of the InterCloud vision is and it's not just one data center, it's a global set of data centers. And it's not just Cisco data centers, but it's data centers that are provided by our partners. So we are able to put InterCloud into our larger partner community of service providers and now take all of the Cisco services and deliver them through that kind of cloud. And so that we're doing this in concert with our partners where before there was concern you would compete with your service providers. We decided, and something I learned from the Sun Microsystems, we were doing Sunclat, you do things with your partners, you actually have a bigger, broader reach. Right, so it's a combination of service providers having facilities but there also will be Cisco facilities and they'll be connected. Yes, I mean we will put the gear there and we will run it and operate it for them and yet they bring it into the customer base. Can you share with us, what is Cisco's, obviously you have a lot of software but when it comes to SaaS applications, what's kind of the breadth of the software? One of the ones I like to talk about is we all know WebEx. So WebEx is obviously one of those SaaS delivered applications around the globe. And Meraki, it's a change in management. So how do you now, what would enable an easier management paradigm for people who are wireless access points? Well Meraki now does that from the cloud. I think we're going to see more and more of those applications move to the cloud for management of ordinary things like your wireless access points, make it easier that way. So OpenStack played a key role in this because we say how, what is the platform that we are going to now, what is the cloud platform we're going to use in all of these data centers? And that was very clear that it was going to be OpenStack because Cisco, time and again we're finding WebEx is going there, we're going to put Meraki there. A lot of the network function virtualization is going on top of OpenStack. Our video properties are going on OpenStack. So OpenStack is becoming a foundational element of Cisco's total strategy. I was at the 30th anniversary of the Mac in Cupertino a couple of months ago because a bunch of my buddies- Did you get a Mac one? A Mac one? I wasn't involved but I didn't get to sit in the VIP because some friends of mine are old enough to be on the Mac team. But it was fun to have kind of a geeky night celebrating kind of the day, talking about shoving 1K, bartering over 1K from MacPaint and hearing the guys talk. But it really kind of highlighted what at that point was the tinkering, which is the homebrew computer club. Now you're seeing that tinkering coming back and with OpenCompute, Raspberry Pi, you're seeing hardware being a big part of it. So I got to ask you that someone who's been there done that from a couple different cycles, you've seen the innovation. You don't have to be someone's uncle to get a motherboard of Intel. You can get it anywhere now. You can get hardware. My interns are working around with a notebook with a little suitcase, which is an OpenStack cluster. And so they have in there six nodes or whatever, running OpenStack. They can open it up. They can walk in. Let me show you what OpenStack looks like. This is a huge new thing. And even at IBM Impact, we were looking at some of the things there. These young developers, they want to hook up a drone, put some hardware on it, you know, GoPro camera. It's a great time to be a developer here. It really is. And so anybody out there who's not coding, you're missing an awful lot of the fun that's happening today. Let's talk about the OpenCompute. What is that? That's a huge thing. That potentially is the next home, bro. What do you think about what's going on with OpenCompute? I think we're seeing the sort of advantage of Open being applied in lots of these different areas. Even in terms of hardware architecture. So can we open source the architecture that we want a computer to be built out of that we can have a lot more input into it? I know it's just that even back at Sun, we were talking about open sourcing or Spark architecture. So each one of these layers, I think can benefit by having more and more people looking at it, more and more contributors going into it. And it does require a changing business model on the part of vendors for how then are they going to continue to either differentiate or provide the value. It's a great time to be a developer. So I got to ask him on a personal note. He had four kids. One's freshman in college. A lot of other parents I talked to have kids. And the question is, oh, I want to get involved in computer science. And one parent, I won't say their name because they're a friend. Oh, I enrolled them into a Java class. I'm like, oh, don't do that. Not that I didn't think it's Java. But there's more fun to do other things. So I want to ask you, what do you see out there that would help kids get addicted to the funness of computer science? I mean, Java class for an eight-year-old, I wouldn't suggest that's my personal opinion. Python they can do actually with that age. I'm surprised at how young kids do when they pick up a programming language. If your kids are in college, I would have them go up to mobile apps. I mean, that's also something that they could actually show to other people. There's a lot of tools out there for doing this. They can share. They can share that. And it would mean something to them immediately. Yeah, instant gratification, something where they can get instant gratification. We are living in that age. Is there any cool things that you've seen that you'd like to point out? No, I'm as interested actually also in the whole idea of open data. So you've got open compute, open data, for example, in San Francisco, the city has made available the data stream for available parking lots, a parking spot. So that means that people now can develop applications for their mobile phones that anybody can use to promote. The OpenStreetMap project was pretty successful as well. That was very open data. I was talking to the founder of Factual, you know that company, Factual? Yeah, very much into the open data. Five of C is dead now. Five of C is going to be... Okay, we're here at the OpenStack Summit. Lou, final question. I want to get you to share your own words to the folks out there. Why is this point in time 2014 so different than other major sea changes or tectonic shifts? Whatever you want to call it. What's so special about right now happening in the tech industry in the world? I'm not so sure it's that different than what we're going to see in the future. I think it's the beginning of a new set of changes that we're seeing. We're changing. Networking has become virtualized and that we're the whole notion of what is a network of change. We've introduced even with Neutron. Yes, it's had its rough spots because it's the first time anyone tried to build a network as a service. So I think it's the beginning of a new set of changes that you're going to see. And people will point back to this time and say, I remember when we just sort of started to see OpenStack have its first real confidence. Remember when Lou Tucker was on theCUBE talking about that? Yeah. Only 4,000 people. Only 4,000 people. Can you imagine that? I think, yeah, new user experience. Lou, great to have you on theCUBE and hear your perspective. Obviously, your experience and the industry has been great. And to share that with us is being involved as vice chairman of the OpenStack Foundation. It's been a great testament. Congratulations. And Cisco's obviously got some new activities going on. I've been following them pretty closely. This is theCUBE here, live in Atlanta with Lou Tucker of Cisco Systems Vice Chair of Foundation. We'll be right back at the short break. Thank you so much.