 Live from Vancouver, Canada, it's theCUBE at OpenStack Summit Vancouver 2015. Brought to you by headline sponsors EMC and jointly by Red Hat and Cisco with additional sponsorship by Brocade and HP. And now your hosts, John Furrier and Stu Miniman. Okay, welcome back everyone. We are live here in Vancouver, British Columbia for OpenStack Summit. Three days of wall-to-wall coverage. We're in day three, this is Silicon Angles. 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 Angles. Joe Miniman, Stu Miniman, chief analyst for Cloud and infrastructure at Wikibon.com. Our next guest is CUBE alum, Lou Tucker, VP, CTO at Cisco, great to see you. And Mark McLaughlin, OpenStack technical director from Red Hat, welcome to theCUBE for the first time. Welcome guys. So we've got some legends on theCUBE. We've got Lou, we've got Mark. You guys are legends in the tech community. So great to have you on, share your insights. First question I got to ask, you know. I mean, it's an oh shit moment here for OpenStack, Lou. You mentioned that was your sound bite from your speech this morning. People, the cloud is here. People who didn't make the bet need to be running fast to make the bet. OpenStack has legs, it's crossed over to path to maturity, not just POCs anymore. What's your take on it? Do you agree and give some insights on where OpenStack is right now and what's going on with customers, people shipping? Yeah, I think the oh shit moment's happening for the people who are just waking up now. And they're competing against companies who had the ah-ha moment like two years ago that they were really going to get on board with this. And so now there's a real catch up that a lot of companies are having to do. And I saw that, that came actually out of looking at a National Association of Broadcasters, NIB conference, where it's all hardware appliances, all moving to virtualized now, systems now. And the companies that started that two years ago have a tremendous advantage and tremendous head start. And that's important from a competitive advantage standpoint because in this cloud era, the lock-in doesn't exist anymore for competitive advantage, the new lock-in is speed. So there's economies of scale with that advantage, so use that. Do you agree? So Lou, there's some in the journalistic world that have said that BCs are fleeing away, there's some in the analyst world that have said that OpenStack's a science project, what are they missing? They're missing the fact that the customers have been looking for something like OpenStack for a long time. Something that is not produced by a single vendor, it doesn't have that kind of lock-in, something that is the rest of the IT community is working together on. So that's where you see both Red Hat and Cisco working with IBM, with HP and others to bring us out. So it's not the typical thing that analysts think of when they think of one faction against another or something, here we are working together. So that's different. They're having a hard time understanding that. They're having a hard time understanding that. That's old school, the factions working against each other now, that's different. But what's changed? Why is it not the factions against each other? Is it more maturation of open source protocols? I mean protocols, I mean like etiquette or? You've been at this alone. What's your take on that? I mean, I think you're absolutely right. What we're seeing here is the convergence of people really just demanding that competitive advantage of agility and that ability to kind of disrupt themselves before they get disrupted by others. And then this kind of combination of we've created this amazing community that's building what people want, what people need to get that competitive advantage. I mean, I've been around open source for a long time. You don't see many communities like the open stack community really the only one with the amount of kind of diversity and the amount of stuff going on similar to open stack is the Linux community. But I think what's happened here is we've managed to draw the corporate world closer to the actual development happening in the project and a lot more real true kind of collaboration between companies and co-optition between companies happening in this project. And it's really something special. Yeah, so Mark, this is the third year we've done theCUBE at the event. Last year we had a lot of debates with people. Is there enough leadership? Do we need a fanatical dictator because that's what you really need in open source and everything like that. There's been a lot of changes in the board, the governance model, talk about the big tent and powered by open stack have been really one of the main talks that we've had this week. From your viewpoint in open source, where are we, what's the big move that happened over the last year and are we in good shape for going forward? Yeah, I actually gave a talk on this yesterday. It was titled like open stack governance bridges and hierarchies and what I was really talking about was how healthy open stack communities, like healthy even 21st century social organizations are kind of bottom up non hierarchical organizations and leadership is through empathy and empowerment and trust and all this kind of thing. And it's really, I'd like to kind of set some principles for how we want to get to as a community because I think a lot of people within this community like hundreds and hundreds and hundreds of developers contributing here who've never really experienced this kind of community before who aren't really aware of this kind of abstract concept of a non hierarchical organization we're trying to get to and I think we do need some more leadership within the open stack community that's really aware of what we're in for. Some will say that's more kumbaya and that doesn't scale although it is healthy dynamic to have that non hierarchical. I think it's the only way you can scale. Yeah, absolutely. I do think, I mean, hierarchies scale by growing the base. You get larger and larger and you still got one person or two people at the top of it or that doesn't scale in terms of attacking new markets or going up to new kinds of opportunities that we see going on in open stack today. Yeah, the other thing too I want to point out is also the growth in the market is pretty significant. So there's plenty of beach head. So growth hides a lot of the backbiting and the fighting because there's so much beach head there's fruit on all the trees, if you will, for everyone to play. So there's a lot of room. In terms of scaling, they are a developer community that the analogy I use is back in the day like 2004, 2005 when Linus was really struggling to keep up with what was going on and they created distributed version control. They've created this kind of bottom up model and I think with some of the scaling challenges we're seeing in the open stack community, we're going to see a similar kind of turning point where we just find something new. It's interesting, you know, we're a student of all of open source, we've given our industry experience to you guys as well. Linux had something that was going on for it around it which was they had hardware manufacturers. You had the client server guys, you had the proprietary Unix guys who were like, let's kill that. So that was the, here you don't have that. You have a lot of scale out, open commodity. So there's no real proprietary targets, kind of elusive to say, that's the bad guy. We mentioned lock-in. How does that change our world? I mean, that makes a little bit more frictionless in a way of scaling up. I think it comes down to what, you know, at the end of the day what a company has to deliver to a customer is some value proposition. And so what we're saying, the value proposition shouldn't be in how your code is different than another person's code when the customers are actually trying to do the same thing. Instead it has to be all of the other associated values around that, that they were supporting it, that they're investing in the architecture, that they're making investment in the community itself, taking the feedback from customers, putting that back through the process. So I think in OpenStack we've tried to create a model that allows all those companies to be successful even though what we're not trying to differentiate on is actually the code itself. It's in all of the everything else around the code. I was talking to Stu before you guys came on about this is probably going to be a Harvard case study on OpenStack because it has nine lies, right? Because it's always had these moments of speed bumps where something could have happened, it could have been a marketing program, then vote with your code and the evolution over the years has been really fantastic to watch the adaptability of the core community that the passionate, I call founders, not real founders, but you know what I'm saying, the early kernel of people. So I got to ask you guys your perspective on what's going on today that I think is super successful in OpenStack is that you get the passion and learning and you got people sitting on floors in the technical sessions and then you got the big whales coming in, you got Oracle, Cisco, IBM, HP, EMC, and so the big guys are coming in but they can't dominate, but they have to ingratiate in. So it's a nice balance between that, we're the European Union. So comment on that, why is that successful? Why is it successful? It creates an amazing community culture that's actually really, really hard to establish when you've got lots of companies involved. Everybody in the OpenStack community, all the developers involved, they leave aside their affiliations and they're really representing the interests of their communities. And like I've read that, we struggle really, or not struggle, we actually push that really hard but our developers really mentor them, teach them how to do that. I think that's where we've created this environment where people can collaborate and get stuff done without actually working for like a... So I'll call that developer-first environment. Yeah, absolutely. So it's not company-first, it's developer-first. And one of the benefits though, the OpenStack has brought this methodology about how you develop code in the open to many, many more developers than probably any other project I know of today. I know within Cisco, I'm constantly bringing traditional Cisco engineers who have never worked in this kind of environment into OpenStack, and they love it. I mean, first of all, because developers, quite frankly, they develop code because they want other people to use it, they want to be recognized for that. And that's what happens in an open environment like this. They get recognition, not just from their own Cisco employees but from people they've never met before, from people that are in competitors and they get to learn and be recognized for at the same time. Yeah, so one of the things that really caught my attention this week is some of the new things that people are doing that they couldn't do before. When OpenStack started, it was like, oh, is it the AWS killer? Oh, does it help people get off VMware? But talk about things like NFV, talk about new applications, new business models talking to Kino. Well, what do you see? What's kind of the real proof points that this is going to help drive IT forward in the industry? You know, we alluded to it earlier, looking at the media landscape, moving from hardened appliances now to virtualized systems. Each new, each industry is now going through that. So NFV is a prime example of that, that many of the traditional network service providers or whatever are really struggling. Their costs are, you know, people are willing to pay less and less for those kinds of services and their costs are remaining the same. So they are turning to virtualization. And when they turn to virtualization as a concept, they're going, where's the platform for that? And so that's when they are saying, OpenStack is a platform. So we have OpenNFV, which is taking multiple open source projects and bringing them together into a solution that the carriers can use. What's really interesting for me around NFV is just how we've gone over the past year from people not understanding what NFV is, how it was all about agility, for telco is being able to move faster in terms of the application they're deploying and how it had any kind of relationship with OpenStack, right? So people at first, when they saw this use case, were kind of like, that's not elastic, OpenStack can never do that. But we've worked through the issues and we've kind of seen how it's whether it's IPv6 or whether it's kind of NUMA awareness on compute nodes or whether it's kind of better networking or more flexible networking support or whatever. All of those particular advancements all make sense in a broader context than just NFV. So it's a community of diverse interests and people are kind of able to apply for it. Where's the hotspots guys? Where's the action at and the hot action and also the action where you need to improve faster? For instance, you're out of service by NFV. A lot of pressure on certainly Neutron and other areas within the service provider, but there's a huge different diverse use case. I mean, service providers have different needs and say an enterprise, let's see the scales there. So what's the difference in finding the common needs? Everybody has a need for a platform that works that can be highly resilient, that can be easy to manage, low cost to operation, operate and everything else. So those are common across almost all these cases. So those are foundational. So the foundational pieces are still a big emphasis of where we have to go. But then I think we are seeing this kind of verticalization. NFV, we're seeing it in video. We're seeing that containers. And so here the classic case we've got, are those competing technologies? Cloud computing versus the container model, whatever. In a monolithic you say, yes, but in a distributed world, the needs of the customer ultimately will trump, right? In fact, in the single organization, you probably have needs for both. And so this is where OpenStack is an open community. So we can actually embrace these multiple differences. I've got to ask you on that point, because you're highlighting the thing we were talking earlier, is that this theme that we're seeing from our report, looking at all the data points and the beautiful views here in Vancouver, is a real emphasis of architecture of this show. So there's not a lot of hype. Last year it was at Atlanta, oh, containers, and they weren't even, I thought it wasn't even around them, but they were kind of driving in. So it kind of hijacked the conversation space. Here the conversation in the hallways is really architectural. Architects are here. So to your point, you have a foundational set of platform, but this undefined top of the solution stack, if you will, for lack of a better description, is developing. So yes, architects are here, but I got to ask you, what language do the customers speak? Because one of the things we're saying is that, they don't say I want a platform as a service. They don't say I want infrastructure as a service. They want services. So what is the language of the customer and how are you guys balancing in OpenStack? How to be at least compatible there? Well, I think you're right in that things have evolved over the last year, especially in terms of how we talk about containers. I think, and how people coming to this event are talking about containers. A year ago, it was the new thing. People hadn't necessarily dug into it too much. They now understand it. They understand the challenges about managing the infrastructure under it. And so they are now talking about the applications that they have and the mixture of, they'll be able to run some services in containers and some more traditional applications will run on VMs and how to connect them all together and how the storage and networking integration will all work. It's getting down to real kind of practical how we're going to bring these two modes together and how we're going to integrate them together for specific workloads. It's all about the workloads, right? So we have a question from on the crowd chat here. So crowdchat.net slash OpenStack is our engagement container. Are we calling it kind of a joke on Docker but it's a whole lot of discussion. So David Deans, he's red with red hats. Walmart's keynote and others, the first day has talked less about infrastructure more about business outcomes, which has kind of become cliche but it does mean something. This is a unique turning point where CEOs find some meaning here. I guess what he's teasing out is business outcome is what people want and that is in the eye of the beholder which is the customer. So OpenStack is a resource set of technologies. It's an enabler for those outcomes and that's where we're seeing even Bank of America was saying they want a software defined infrastructure and they define that a couple of things that are matter to them. Agility, speed of new deployments. They need to roll out new services to their branch offices or something that's a pop-up store because they may want to do banking or whatever in a matter of months and so they are moving to cloud to get that kind of agility. Agility is an outcome that a business that can move faster can ward off the disruptors of their business. They want it at lower cost so they need it fully automated. We had Fias Honor from Cisco earlier, you're going back, oh in 1992 when I joined Cisco and I'm like we're kind of having a historical kind of throwback Wednesday when I call it. I said, so I said, okay Cisco made a lot of money with TCP IP as a disruptive enabler and that created internet working and the buzzword now is intercloud or interclouding I'm calling it. But that was an enabler, very disruptive and what happened on top of that was a lot of wealth creation, great outcomes for customers. Cisco became a huge company. What is that today? You mentioned, is it agile? Is it cloud? What is the disruptive enabler? Similar in analog, that's something directly but what TCPI did was enabled in industry, right? OpenStack in the same vein is enabling an industry transformation not just customer outcomes, get vendors, you've got developers. So what's your vision on that guys? Because share, what is the enabler? Is it just being agile? Is it just being faster? Is that the new differentiation? All these things kind of dovetail off that. So for me just preparing for my talk yesterday I've gone back over the agile manifesto and for me that's kind of where this all starts, right? So agile was all about software development and kind of smaller teams and more engaged with business value but it's moved on now to DevOps and how to deliver continuous value to customers and it's about those business outcomes and it's about how to operate your applications and in order to do that you need this automatable infrastructure under that. So to me it's all this kind of continuum of what came from the agile manifesto and what's led there in terms of DevOps and so it's all about new ways of developing software and new ways of bringing software to the market and I think that's what's driving them. It's freedom, it's unshackling innovation, right? So that's essentially enabling, right? And it's getting out of the way of developers. What we've always seen in the past that IT often has been an inhibitor towards application development that you have to trial. It's a trouble ticket to get a server deployed or to get that port open there's something like that. More and more we are automating these systems and they're becoming self-service. Self-service I think is still the strongest. And programmable too. And so it has to be programmable because it's not just through a GUI or something you need to be able to have other systems. Once you make programmable systems start getting tied together. So one system can start to drive the other and you can take through automation the cost of delivering that. That's the lever right there. So I think it used to be about cost now it's about speed. And I think speed Trump's cost every time. If you're competitors out there with a competing offer faster than you are that's serious. Talk about the history of open source. I'll say you know going back to the early days you know the shoulders of giants is what everyone says. But right now we're in an interesting modernization of open source. It's tier one. It's first class citizen. You know surely Red Hat being made their bones by coming as an alternative. Now they're running 10 year SLAs the open stack for three years. Cisco's now in the mix with open stack. What is the open source model for today? It's really interesting and cool right now where you have big companies like EMC transforming their entire operation to be open source. I mean that's where it's due. They never would have done that 10 years ago. So there's a business model now for corporate America and you mentioned earlier the factions. I remember the IETF and IEEE, these W3C all these different committees would be fighting. I think it's the way customers are now demanding that their vendors behave. They are saying I no longer want you fighting and all that will you guys just get together. We used to do it through standards bodies. Get together and figure out the standard because and then we'll buy your wares. Now they're saying we need the implementations of that. So get together so it's community driven software that I think we're using open source as a vehicle to achieve that goal. It goes beyond community. It's actually execution in the open. It is. Absolutely and it's about people realizing what you can achieve through collaboration. But the really interesting thing is it's actually really hard to achieve that kind of community that is really collaborative. It's very easy to push a code out there on good job. It's very easy to create a foundation around your project. It's very easy to put yourself out there but actually to engage people and build that collaborative environment. Open stack is done. But also now customers actually look at the leaderboard so to speak and look at what's been closed. Oh that house has been on the market for two months. Oh it must not be good. With code it's like that's an open item, open issue, hasn't been closed. It's a shit community. It's got nothing going on. I'm not buying this. So customers can actually now audit the progress. Exactly right. So that transparency that's always been associated with open source I think is one of the other things. Yeah so one of the big challenges is it's great when everybody's contributing and everybody's trying to figure out where it's go when the money starts coming in. How do we keep that going? We've had lots of debates as to how many successful billion dollar open source companies are? Well Mark you're working for the one. So when we asked Jim Whitehurst last year why aren't there more? He said because selling free is really tough. Yeah it's really hard. I talked about that in my keynote yesterday. It's like if the software is free why do people pay red hat? And you know the answer I used yesterday was expertise in technical sport. But it's a much broader answer than that. But just in the context of OpenStack I think what we've done that's really interesting is we have brought the money closer to the community but we've centered the money around an OpenStack foundation with a mission to protect, empower and promote the project and not to control the project right? We're really trying to empower the project as opposed to kind of control the direction. And those are the principles that we're looking at. That's why it is the model that we have around how do we form these projects and everything else. But back to your question in terms of the commercialization. There's still an evolution here. We are now, I think the next couple of years what we are all looking for is the whole rest of the ecosystem to get developed out. And they are building then their own solutions on top of OpenStack. Because they have OpenStack it makes it lower cost for them to develop the solutions and to move even faster. And then they will start to compete with each other. Yeah I mean Lou, you can talk a little bit about inside the company, I mean I know you're such a huge proponent of open source and what's going on. Talk about how that shift happens and pushes through because the margins, the sales force, the rate of change is just very different than what the traditional Cisco hierarchy is designed for. Well Cisco more and more has, I mean you've heard John Chambers have been talking about it. We are becoming an IT company which means that we really are selling solutions. The IT, the business outcomes. That's what customers are going to get from Cisco. And so open source is an accelerant for that. Because now we don't have to develop everything. We actually can partner with all of the other companies to develop this software. And therefore go to market faster with our solutions. Also security, some people are also arguing that it's more secure in the open because more eyes are on it versus having some proprietary piece of software. Except for a little bit of open SSL out there. So actually I was talking to Charlie and Collin the other day actually on debate on open source and proprietary. And what we're saying is actually being open is not enough, it's how many inspectors you have. So the reason why that bug persisted so long is that we didn't have enough people looking at it. So you can have it open but then you have to have people looking at it as well. And that's something that we always urge everybody in open stack to be doing that. Reviews count are a very important part of that. We need more eyes looking at this. Otherwise the security issues could remain there forever. Yeah so another real challenge out there is when people install something they tend not to change it. A typical Cisco catalyst, you know, you put the code in, you don't change it. Even on open stack, one of the things that we need to bridge is how do I upgrade from one version to the other non-destructively? You don't need to do that. Do you think we can get to a model where it's more online, it's upgraded? If I go use Amazon or I'm using Office 365 I don't think about my versioning. Absolutely. I think we're making huge advances there and I think it's all about creating kind of an infrastructure for deploying and managing open stacks that's automatable so that you can automate the upgrades. One of the big challenges for handling upgrades for a vendor like Red Hat is our customers can deploy open stack in hundreds of different configurations and in order to support the upgrades of all of those it's quite challenging. So I think we'll be taking an approach of really identifying key use cases and automating the upgrades of those but it's a big challenge for open stack. As we're getting close on time I want to get one question and I almost do might have another one but I want to get your perspective as leaders. I lose certainly your reputation, you've been in a lot of experience. Mark we're going to get your take on as well. As for the technical leads out, you mentioned scale, this thing's going to grow like a weed, I mean open stack is horizontally scaling across industries and obviously scaling up in terms of agile functionality. Being a technical lead for the folks that are coming in what is your advice and what's your vision and etiquette or what's the protocols if you will to be a great lead, to be great technical lead. Is it just herding cats? Is it publishing? Is it doing stuff? What's the norms that are forming in the open stack community that you can share with folks that might be watching and saying hey, you know, you could be a participant and if you want to be a technical lead or take a project and drive it home what are the great use cases, success factors you've seen and share your insight and experience? Well, the advice I give to people is number one find your niche, find some way that you're passionate about that you can find a way to contribute and start from there. Really empower yourself to actually, you know, you see a way forward you can get it done and then start empowering other people. Start setting an example for other people and start showing them how to kind of achieve what you've achieved. Through code or through... Oh, it's getting stuff done. That's the only currency here. So we run as a meritocracy. So developers, you know, elect their leaders so that the leaders who are going to be elected are going to be those people that the developers felt most comfortable with leading them. And so this is what we're hoping to accomplish here. So I would suggest, you know, it really starts from what are you doing for somebody else before you're trying to drive your own agenda. It takes that kind of selfless. It's a social equation. A leader is always selfless. It's always for the team's success, not their own. It's a social equation. I mean, it's just common sense in a way, right? You know, be a cool person. Can help people. That's right. Contribute code. Post something on the board. Exactly right. That's how it comes back. And I think that therefore the, I've always stressed with my folks that we start to introduce people in open tech, start with reviews. Start doing something for somebody else, pushing other people's code through the system. Then, guess what? You have an instant audience when we now want to go forward with something, it'll be deep level. It creates currency and self-governance currency as well. And that flywheel just is a- It builds relationships and all of this is about relationships and the interactions between individuals. Yeah, I mean, and the fruit that comes off the tree, everyone shares, right? So that is the fruits of the labor. That's the way we operate on the board itself. Come into, and there's a lot of decisions that we have to make. And it's a rather large board and everything else. And we were there to support each other because we have a shared goal and a shared mission that we're after. And it's a lot less about our individual company's agendas, which I think a lot of times people view it that way. There's old benders involved in this. No, we were there because we actually have the experience and we were there to drive the community forward. So Cisco Open was a hashtag I saw or open Cisco or Cisco Open was a hashtag I saw trending on our CrowdChat dashboard. So what does that mean? Is Cisco Open? And is that internally the momentum? And are you winning? Are you winning that fight? Is it well received? Share some insight. I think it's trying to get across there is a new Cisco in terms of how we're looking at software. And that open source is a key part of our strategy going forward, so we wanted to say it with having such a hashtag. Yeah, great, we love that hashtag. Yeah, so the last question I have, since you're both on the board, is love what to kind of the big tent and really the certification, but there's a balance between how do we get it working and how do we keep innovating? And I heard from a few people that said, hey, devs want to work on stuff that's important. And if it's not core, are they going to still be incented to do that? So how do we keep innovating and stabilizing? Those are usually pulling in two different directions. Yeah, I think there's, it's a community of diverse interests. So some people are going to be really passionate, really focused on improving the core and really tackling those kind of really critical use cases to open stack. But there's always going to be people who want to experiment with different models and there's new projects popping up the whole time. I think we're managing to kind of thread that needle and balance between the two really well in open stack. There's no shortage of people who want to work on the core projects and no shortage of people who want to work on innovations. Great to have your insights on theCUBE. Appreciate it, we're stuck for time now. I want to give you guys the final word. For the folks who aren't here, obviously Vancouver, beautiful city, clear skies, no rains, beautiful weather, what's happening here at this show? What's the vibe? Share some color around what's going on at the physical event, your perspectives, the hallway conversations, what's the vibe? Books not here, give them a taste. As you can see, it's tremendously exciting. I, in many ways, keep expecting things to sort of start to level off a bit. If anything, though, the excitement and the participation is just continuing to grow. Yeah, and it's scaling out a lot. It's more exciting in more and more rooms. Yep, yep. There are all we're growing, we can never seem to, you know, we try to figure out how many people in each of these sessions, in each of the sessions have been totally packed. Yeah, I mean, hopefully, I mean, it's ridiculously amazing. Look at where all the videos are up, they're up within hours after they're being presented here. So even if you can't make it here, you can participate. And that's what I'll be doing when I go back to San Francisco. I was even doing it myself last night trying to catch up on some of the talks and the number of talks and the different topics. We got our crowd show, we got our crowd page at theCUBE. We're bringing all the data to you. Guys, Lou and Mark, thanks for coming on theCUBE, really appreciate it. We'll be back right back after this short break. Live in British Columbia, in Vancouver, for Pistach Summit, third day, wall-to-wall covers Silicon Angle. We'll be right back after this short break. Thank you. Thank you.