 Okay, we're back here live at the OpenStack Enterprise Forum, O-E forums, the hashtag. Go to crowdchat.net slash O-E forum. We're opening up a public timeline across all the social networks, our new CrowdChat beta preview. We're going to leave it open. We already have 2.2 million timelines reached and over 200 comments. So go there, comment, thread. It's like a Reddit for social. And we're taking all community comments. Thanks to Cube, I'm John Furrier with Dave Vellante and our next guest is Raj Dutt's SVP of Technology at InterNAP, welcome to theCUBE. Thanks, thanks for having me, John. So pack house here physically at the event in Mountain View, Silicon Valley, online a lot of activity on Twitter. Enterprise is hot, that's where the action is. And people want to know why. And finally the guy raises hand in the middle of the panel. I've been here two hours and someone explained to me what is OpenStack on the Enterprise breaking out really mean? I think you gave the most candid answer. So why don't you tell the folks out there what is OpenStack really mean in the Enterprise? So to me it means a realization that look, it's 2014, Open Source is going to win. Open Standards are going to win. VMware is a dinosaur. And you're going to see a massive commoditization and OpenStack is a vehicle that the community and the global development community is going to be able to knock off billions of dollars off of market cap off VMware. And I don't mean to be provocative about it but we really believe that Open Source will win. So Mark McCauley online said the following. What happens when we commoditize apps running on open commoditized hardware and open commoditized OS and unlocked open commoditized endpoints? Question, how do we make money putting ourselves out of business? Of course I had a comment but I want to get your take. What's your answer to him? So my answer to that I guess is look you can commoditize your way all the way down the stack from cloud orchestration to the hardware to the data center to the network to the fiber in the ground to the way that you route bits on the internet. But at the end of the day, companies can still differentiate on the value proposition of their application. I don't think that's ever going to get commoditized. And as an application developer, you want to be able to focus on making your app awesome without worrying about all this auxiliary stuff. So your comment about VMware being a dinosaur presumably you're talking about it going away at some point, getting disrupted at some point but of course dinosaurs stayed around for many, many, many years. You would agree I presume that it's very functional. And there's a lot of stuff out there. Don't get me wrong. VMware is a fantastic, very stable product. They've made their name in terms of enabling the enterprise but that was the first wave. We're now in the second wave. So your point about being a dinosaur is that because it's not open source, right? You're talking about from a business model standpoint, correct? I just wanted to. Yeah, no, fundamentally the fact that they're not open source is a huge disadvantage. Yeah, so now of course they'll play the game, right? Pieces are open source. Oh, hey, you can have our SDN or the NICERA piece because nobody's buying it anyway so we're going to put it into the open source. Smart move though. Pat Gelsinger's no fool. And so the trick for them is to try to stay ahead functionally and the trick for OpenStack is to try to close that gap as fast as possible. So what's your prognosis for that? My prognosis is while I have a lot of respect for VMware I just don't see how they can overcome this huge tidal wave that is a global open source collaborative effort. I mean, it's happened again and again, whether you want to look at Linux, you want to look at Apache, the same story is going to play out here. And I think that they can reorient themselves to stay relevant. But at the end of the day, they've been a huge enabler in server consolidation. They have done almost nothing in terms of agility. So server consolidation, enterprises have servers running at 20% capacity. I'm going to consolidate a heterogeneous environment. But you tell me what public cloud is running VMware at scale? What internet-centric companies doing that? Let's drill down. That scale is a big gap for those guys. They don't have it right now. Well, that's the point. He preached a good point. The server consolidation works great in the down market. You want to cut costs. Now we're in a growth market. Some say bubble, but obviously growth. New apps, connected endpoints, all that stuff. Okay, growth market. Okay, hyperscale. The hyperscale guys are kick-ass dudes. They're eating glass, spitting out nails. Go to an enterprise they've been consolidated down to, I'm the DBM, the storage admin. Sure. And the enterprises are doing. And now you're going to be an agile DevOps guy? Not really. So cloud ops seems to be the kind of halfway point, halfway house, kind of the recovery zone for want to be DevOps guys. But isn't that what Pivotal is all about? I mean, you know, think about Tucci for a minute, right? EMC is spinning off Pivotal. Kind of an interesting move, right? But that's PowerPoint. More open source mojo, right? Sure. Yeah, but it's one level removed from, you know, what we're talking about here from a cloud orchestration standpoint. Yeah, that's really a pass lawyer. It's more agility. Yeah. Sure. Now, this is a great segment. Staying on the controversial comments, Randy Baez said in the crowd chat, he said there's no place for OpenStack in the public cloud. 80% of the deployments are private cloud. There's no place in the public cloud. I presume you disagree with that. Yeah, no, we definitely disagree. Why? What gives you confidence that he's wrong? Look, okay. So Amazon's the 800 pound gorilla, right? So what is, but it's still early days in this whole transition, right? People act like it's game over. I'm of the opinion that the game's just beginning. So OpenStack is the main answer that any new player or up and coming player has to defend against Amazon. Look at the very history of the project. OpenStack exists because it was a defensive move by Rackspace acknowledging that they were going to have a very difficult time catching up with Amazon. And hiring people that had cork on them. So I buy that, I buy that, but that's a sell side view. What about the customer view? The customer view, okay, I will be the first person to admit that today's customers don't necessarily care that they are using an OpenStack public cloud. That's not part of their buying decision. But given a few years, interoperability is going to start to matter. Portability is going to start to matter. We're still in early days. Once that starts to happen, I think people, customers and users are going to start demanding OpenStack in the same way they've demanded Linux, Apache, so on and so forth. I mean, you would think that the days of the DeGiori standard are dead, but here's Amazon with the, you know. Hey, look at what Chris Kemp said at this conference. I thought it was a very popular statement that Amazon would eventually end up supporting the OpenStack. And I thought it was tongue in cheek, but he's going to come on, so we're going to have to ask him. But Dave, here's the thing. We started OpenCompute this week. Building large public clouds is fricking difficult. Even Amazon and Facebook with, that's why they went OpenCompute, right? So it is really fricking hard. But that's why I tweeted out. I said, I think eventually the OpenStack's going to be very appealing to public cloud providers. Hold on, hold on. Let me fix my thoughts. Okay, that's a given. We talked about that. I don't think we need to argue that. It's really fricking hard. You know that. But now, Enterprise is one hyperscale. So their demand is for hyperscale. But can they operate? So the question is, whoever can put hyperscale into a simplistic package with automation? I think that's where it's the OpenStack. And that's why I think it's hard to do both a public cloud. That's the opportunity in the challenge right now. That's why Amazon told us that they don't go OpenCompute because quite frankly, they're already doing all this and OpenCompute does this. So as public cloud providers, they got custom, basically it's a custom project. Yeah, look, at the end of the day, servers or servers, everyone's using the same x86 commodity server. It doesn't matter whether you're getting a Dell, an HP, Quanta, OpenCompute, NonOpen, OpenCompute, SuperMicro. The price difference is actually very, very, I mean, across all vendors, it's commoditizing. And as a public cloud provider, do we really care if we're using Dell, or HP, or SuperMicro? It's a supply chain issue. It's really a supply chain issue, right? It's optimizing for the supply chain. And it's about sharing the institutional knowledge that has been gathered at the hardware level, at the data center build level, just like at the software level, right? It's a realization that, hey, everyone's solving the same problem over and over again, and let's share what we're doing in server design and data center design. But again, at scale, that means billions in a supply chain. People who need support, Dave, we're not going to call Facebook for OCP. They have to do it themselves. The people who build their own clouds have to basically say, hey, I'm going to optimize for cost effectiveness and scale. OCP needs solutions. We talked about that. All right, guys, I have to ask you another question. I'm sorry about the rapid fire approach here. We usually like to savor the cube, but we're under the gun here. Explain why bare metal is cheaper. It's somewhat non-intuitive to me. You would think if I'm sharing the infrastructure, shouldn't it be less expensive? Explain what you meant. So, okay, look, at the end of the day, I have a server. I cut it up into 20 pieces and virtualize it. My total revenue for that server as a service provider is going to be better than what I'm going to get if I just sell it as a dedicated server, right? So there's a margin question here. Right. So, I mean, at the end of the day, would I rather as an app sell 100% virtualized servers? Sure, but you know what? Look at companies like SoftLayer, Rackspace, whoever you want to look at. They've built their entire business on bare metal. And how else are you going to ensure to a customer that they're not going to have a noisy neighbor? They're going to have no co-tenancy issues. They're going to have fast access to local storage. I don't know any other way to do it. So this is the debate that we've had with Amazon. And you know, you kind of have a debate with Andy Jassy, but it's hard to debate Andy Jassy. But at any rate, that apples to apples, it's more expensive to use a multi-tenant approach than it is to do bare metal because of the complexity. I think you're looking at a margin difference between the two worlds. Yeah, oh, so from a supply standpoint, again, the margin's great, like you said, from a supplier standpoint, but in fact, even from a customer standpoint to build an apples to apples, true enterprise infrastructure on the public cloud with all the security, compliance, everything else is more expensive. It's always been our premise. I think that's what you're saying, right? Well, we are, but at the end of the day, look, bare metal is not that different. I mean, at the end of the day, you're providing a server compute capacity to the customer. Does that customer really care whether it's bare metal or virtualized? If it's provided over an API, it's elastic, they can throw it away when they're done. But the customer prefer that it's bare metal, I would presume. And so that's what we found. We found we have had many, many customers who've come from Amazon, they get to a certain dollar amount spend, and then they realize that it's getting out of control. They're on the wrong side of the cost curve and they get to the point where they can basically reboot their whole application on bare metal, save money and get better performance. Rental is almost always more expensive than owning. Assuming you can utilize the- There enough, which is why we offer co-location. And in spite of what people say, co-location is not dead and is a valid offering for many, many companies. Well, sure, right, absolutely. Steve Rajbal, hey, great to have you on theCUBE. We definitely want to get you on our roll of decks list because we've got a lot of content there, you know, hot area, hybrid and- Oh, before you let him go, one more question, go ahead. So I wanted to ask you about quality of service. Are you able to monetize quality of service on the storage side, the elastic block storage analog in your world? We're starting to. Today we're not as effective as we'd like in monetizing it, but in our next generation cloud, which is currently in beta and goes into deployment at first half of this year, we will absolutely monetize and bill on metrics like IOPS, megabits per second, allow our customers to build an exact profile of what they need for their storage. I mean, it's early days, I really don't see anybody doing that in a sophisticated way, but the technology's here. It's solving the problems that have existed, performance without compromise. All right, thank you, John. Thank you. Okay, no, thanks a lot, guys, great content. I wish we had more time. Normally we like to do a little bit longer, but we're under a tight window here. Raj, thanks for coming on theCUBE. You're a jack and athlete, love the action. This is theCUBE, we'll be right back with our next guest after this short break. I'm John with Dave here inside of the OpenStack Enterprise Forum, crowdchat.net slash OE Forum. OpenChat across all networks, go there for the public timeline.