 Okay, we're back at VMworld 2011 for day four. My name is John Furrier with SiliconANGLE.tv and SiliconANGLE.com, and I'm here with my co-host. I'm Dave Vellante from wikibon.org. This is day four, John. Pretty amazing, wall-to-wall coverage. And we're here with Paul Frutan, who is the CTO of Nervonix, the cloud storage innovator. Paul, former Google exec. Welcome. Thank you, Chris Beer. So you're at Google. Google, obviously, is in the forefront of everything that they've been dominating the world for the past decade in terms of search, but data centers and infrastructure, they're massive. Facebook is now nipping at their heels, massive scale. The question I have is, you were responsible for a ton of stuff. Tell us at Google, what did you do there? I mean, what was the scope? Well, my org was responsible for operating all the production servers. So all the machines that we had around the world that served the various Google applications and all the backend stuff, we put them up, took care of them while they were alive and then took them down, took care of them later on when they were done. And you guys grew like crazy. Over the course of your career at Google, how many data centers did you guys build out? We announced five or six data centers, I believe, while I was there. Very, very large data centers. I mean, these are multiple hundreds of million dollars a piece, big scale. So we covered the news. When you left Google and went to Nervonix, really fast, hot growing startup in cloud storage. So cloud storage is all the rage. Why Nervonix? I mean, so what attracted you to Nervonix? Well, you know, Google's full of a lot of smart people. And if you spend a little bit of time there, you can learn a lot. And I like the things that I did learn a lot about big infrastructure, big scale and how it all works together. Saw a company like Nervonix, seen them for a couple of years, I followed their moves and I just thought they had a very unique product. They've been around business for a while. So, and they have some good customers. And for someone like me, I discovered that I'm actually more passionate about kind of a smaller startup environment. And Nervonix seemed like a great fit for making a difference. So we've been on theCUBE here this our fourth day. theCUBE is our flagship telecast. We go out and talk to everyone, Pat Gelsinger, all the top executives here at VMware, around the industry, bloggers, VCs, and everyone's all jazz about cloud. And Nervonix is obviously in the cloud business, with cloud storage. But the thread that's been kind of going through theCUBE here is that this is the year of winners and losers, okay? The demand is high for products. So what's your take on the cloud space, winners and losers? And what's going to be the key kind of variable that makes the winners kind of tip over? So the way I look at it, it seems like people are finally understanding and realizing what is cloud and what isn't. You can't take a boss traditionally. He's been around for 20 years and we sell it as X, Y, or Z, put a cloud label on it and call the cloud. And that's been going off for some time now. It's the hot thing everyone wants to jump on. But then there are some actual cloud products out there. And finally, after so much time, people are understanding which are real and which ones are not. And the ones that are real and have a good product will stay. The other ones will either go where, go back to their traditional markets. Paul, what are the storage implications of true cloud and machines at scale? The storage implications of this true cloud, this real cloud and machines at scale? So probably the easiest, the most impactful thing of it is the end user doesn't have to worry about it. That's what cloud's supposed to be, is the end user puts the data in and forgets about it. It's to all handle inside the cloud, the replication is, DR is, having number of different copies of the data, ensuring data integrity, ensuring data availability is all handled by the cloud. And the user shouldn't have to worry about it. That's what the real piece is. You shouldn't have to go and buy autonomous machines, figure out how to put it together, how to architect the entire infrastructure. This is all supposed to be handled by the cloud. Now at Google, you had a bunch of PhDs running around to figure all that stuff out. You have to, first of all, what can you learn from that that you can bring to Nirvonix and how do you replicate that? Because you don't have an army of PhDs, I'm sure you have some, not nearly as many. So how do you take what you did at Google and that concept and enterpriseize it, if you will, and package it? Sure, so the Google model's a little bit different in the sense that at scale, even some of the largest companies aren't anywhere near that size of scale. What's nice about having worked in that environment is you see all the problems that you are likely to encounter in the next 10 years as storage and everything grows. And once you know what to encounter, you can take the right steps now to make sure that those are addressed when they come along. In terms of the enterprise, I mean we can bring that, we can look at all the positive aspects of it and how you do it without 100 PhDs by focusing on one piece, right? Nirvonix does cloud storage, that's what I like about it. Not compute, not a bunch of different products. We do cloud storage, that's all we do, and that's how we can be good at it. One of the things that we've been hearing again in theCUBE this week, and this is the best part of going on day four, where you get to get all the data from the marketplace, but we've been talking, Dave and I have been talking about the demand right now is so strong for cloud that it kind of puts in check the hype, right? So in Google obviously when you were at Google, there was no debate, they were growing fast. So there was no like pie in the sky plans. They had to deliver value in building out the data center. So talk about what you see at Nirvonix, because you guys have real customers, you guys are delivering cloud storage. What is your plan at Nirvonix on your build out? What's your strategy and kind of what are you thinking? So we're looking at all the important aspects of the cloud, being the incremental growth ability, the pay for what you use, having the replication, the access and data integrity, and looking at how can we continue to bring it to the customers more efficiently, faster and cheaper, right? That's where we need to take this, is to make all the great benefits available to them at a reasonable cost, at the right performance that they need. And we actually have a couple of interesting things that are coming up soon, and they'll really make a difference in the market. We already are the only company that will do private, public and a hybrid cloud in the real sense that we can actually bring the cloud to you. We make you part of our cloud in the hybrid sense. And that's a very key piece of product that we have, and we're going to expand it and do more with it. So following up on that, Dougie, I mean, I think a cloud, I think a cloud is cheap and deep. And you're semi-cheap and deep. So is that right? And how do you add value beyond that just, okay, I want to put it in S3 or just put it out there in the cloud? Sure. And you touched upon that a little bit with the public and private. I wonder if you could expand on that. All right, so one of the things that will soon start discovering, the market will soon discover, start discovering that not all the clouds are the same. Right, we talk about storage and it's kind of all thrown in the bucket, but it's not all the same. And then it's not just about chiefs. There are actual different features that different clouds have. For example, our cloud has data consistency, none of the other ones do. S3 doesn't, OpenStack doesn't, others don't. Just as an example of a feature that we have that they don't. The hybrid piece is not a piece that we offer that they don't. So it is not just about that. So you take those things and then you combine it with enterprise-level support. We have phone numbers, we have texts that will answer your call. If there's problems, we can address it. If there's customization to be done, we can do that. If there are a number of partners, many of them here, that needs to be worked with these guys to make the entire user experience better. And we do all those things, whereas a lot of these other ones don't. Okay, so you can get a premium for that and presumably make a profit at that. And who do you, I mean, I don't really see anybody else going hard after that business. I presume it's kind of a traditional service providers. Maybe it's IBM and HP and maybe Oracle eventually, but there's really nobody else that's specializing in that that I see. Is that right or why is that? It certainly seems that way. Why is it, well, the simplest answer is because it's hard. You know, if it was easy, they would be doing it, but it's not. I think a lot of these companies are looking at trying to do something like that. And as soon as they get down the road, they realize it's not that easy. You just can't take six boxes, stick them together and call the cloud. As soon as you get your first customer they start asking you questions about billing and shared tendency and all that stuff, you only run into a problem. So we're actually now working with a lot of these partners to help them set up these clouds either for their internal use or for them to use for their customers. So EMC has kind of an interesting example. I wonder if we could talk about that. I mean, essentially they got Atmos, which is, you know, it's a box, you install it, you pay upfront for it, but they partner with cloud service providers. It's an object store. They essentially, their messaging is very similar in terms of scale and the number of objects you can support and so forth. So can they compete with that sort of virtual integration with their cloud service providers? Do you see that as sort of your biggest competitor? It is an interesting product. It's a good product for some uses. The biggest, one of the biggest products, like you just mentioned, first you have to go buy this huge box, which you may or may not use. You don't know how you will use it because you don't upfront know how much fast storage you need, how much slow storage you need. There's a cost issue. And then the global object store, right? In our cloud, you put something in one location. It is the same object everywhere around the world all the time. And from what I know from Atmos, that's not the case. It is in pieces that you have the Atmos box, but globally, anywhere in the world, from any connection on the internet, it's not all the same. Well, I got to ask you because it's one of my favorite terms on the planet right now is, well for a while, cloud washing, right? So we've been talking about, obviously on the reality of the marketplace, there's demand. There's real people spending money and investing dollars for real products. And you got new initiatives like OpenStack, Citrix just bought cloud.com, which has been doing some good work. So you have people racing to the market with solutions. So two things, talk about cloud washing. Who's out there cloud washing in your opinion? And two, talk about the viability of, say, OpenStack, for example. Sure. So, you know, I don't want to start naming names cloud washing because we're in this area and I might get beat up before I walk out of the door here. They're all here. You know, I think, a lot of them are out there, right? These are companies that are taking the same products that have existed for some time now, which are all some kind of... Hold on, let me ask you differently. How does someone smell test cloud washing? How do you see, what's cloud washing? How do you spot it? What's an indicator? If the exact same product's been around for 20 years, it's probably not cloud. If you don't, if you have to buy a whole bunch of stuff up front and set it up yourself, it's probably not a cloud. If you don't have a global consistent object store, it's probably not a cloud. So, those are all, you know, simple tests that we can see if it's cloud or not. Okay, so talk about OpenStack. They have a lot of momentum right now and they're getting a lot of developers jumping on board. And developers want trust. They want to make sure that they have an environment. And I like OpenStack. I love the initiative concept. Kind of reminds me of the client server days when, you know, all these new standards, kind of, you know, groups came together. But the question is, can they deliver fast enough? And what's your angle on OpenStack? I like OpenStack too. I think, again, it's an interesting product. I think that they did a good job of open sourcing it and getting the community involved. The OpenStack has both the computer and the storage side and they're not the same product. They came from different places and they're at different stages of their development. I should think probably their compute side is probably further ahead than their storage side. But it's not all the same. We talked earlier about the different features. OpenStack is different than what we do. It will work for some customers. Just like S3 will work for some customers. It won't work for some customers. And there are specific features of Nervonix that, like the global object store, the consistency that actually appealed more or worked better in the enterprise environment plus the service side, right? Rackspace, who sponsors OpenStack, would be the first one to tell you about how important service is. And something like Nervonix, RCloud, we stand behind it. There's a company that will support you. You deploy an open source project, much like any other open source project, you're pretty much on your own to figure it out unless you can get some of the community to happen to like the feature that you're looking for. Can you talk to the specific technical limitations of the storage, the object store of OpenStack? Have you taken a close enough look at that at this point? I haven't gotten way deep into it. OpenStack, like most of the other largest computer file system, it has eventual consistency, for example. They have a file size limit. I think that's still around five gigabytes. So just to take that as a simple problem, many of our large customers have much bigger files than that. And now it'll be a couple of different files. What are you guys doing for file size? Do you guys have Nervonix? Do you guys have a lot of... It's unlimited. So what does it... Break it down. What does that mean for a customer? Well, so you take a large customer that may want to store movies, right? Original movies or whatever in the cloud. These are multi, multiple hundreds of gigabytes, right? And they want to push it out the cloud. Once they get it on the cloud, then it exists in a number of different places. The DR is taking care of it. They can access it from anywhere and they can use it to distribute to other studios, customers, etc. that they have. That's not possible when you have to do small chunks. You have to figure out how you're going to divide it up. You have to figure out how do you recombine it. And when it's stored, how do you ensure the integrity of each one of those pieces? So my question is, you mentioned there's a lot of people potentially here cloud washing, but this VMworld and VMware ecosystems developing, right? So virtualization is now evolving. What are you guys doing here at VMworld and with virtualization? How are you guys using it in your build plan? So our product, I think, works very well with VMware, with products like Veeam and Commvault and other backup because we can take snapshots, for example, virtual machine disks that exist. You can push that up to the cloud, both for DR purposes, for backups, and also for distribution. You have a virtual machine image that you may want to distribute to a number of your offices. You put that up into cloud, they go pick it up, and you always have a copy there available to you in case something happens. Or for example, if you want to do troubleshooting, you take that, you push it up to the cloud, someone else picks it up, looks at it, and they do whatever they need to, they push it back into the cloud to distribute. So it's all there. So you're CEO, Scott Chenero, who was on the Cube on a panel, mentioned you guys have hundreds of clients, customers. What are the main things that they're doing in the cloud with you guys? What's the, you know, what are the core apps? Why are they going to you guys? Because that's a pretty impressive number. I mean, not a lot of people can say, both, we're real cloud and we have hundreds and hundreds of customers. We have an array of the different uses that customers have. They are using for DR purposes, they're using it for backup, they're using it for archiving, they're using it for distribution of data. So we have a number of different customers and so I can't tell you exactly one big application of it, but all of these things are being used. It's all the above. Okay, final question, we're getting the one minute hook here. Obviously at Google, you have, working with a ton of smart people, they really pioneered modern data centers. So that makes a good cloud. What lessons did you learn from Google that you're going to take to Nervonix? And what things do you see coming in forward to you that might be cool things to get your arms around? So probably the most important things I learned is hire very smart people and don't compromise and go big, right? Make things big and reach for disguise because if you try, you might actually get those things done. Okay. All right, Paul Frutan, CTO of Nervonix. Thanks for coming on theCUBE. Great guest. Love to have you back sometime. Thank you. Great, absolutely. And stay right there, everybody. We're going to actually keep it going and thanks again, Paul. We've got another.