 The Cube at EMC World 2014 is brought to you by EMC redefine VCE, innovating the world's first converged infrastructure solution for private cloud computing. Brocade, say goodbye to the status quo and hello to Brocade. Okay, welcome back everyone here live in Las Vegas for EMC World 2014. This is The Cube, this is our flagship program. We go out to the events, extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE, joining my co-host Dave Vellante, co-founder of wikibon.org. And our next guest is Boaz Palaghi, VP and GM of EMC's Advanced Software Division, formerly the CEO and co-founder of Scale.io. Welcome to The Cube. Thank you. So, you're working on stuff, a category, and all of a sudden it's explode as the hottest thing. As Jeremy Burton says, you can't ignore fashion. You learned that trick from Larry Ellison. The stuff that you guys have been working on might not have been fashionable, all of a sudden now it's in fashion. It's like a new suit that everyone wants to wear. So, how do you feel about that? It's a great feeling to start something from scratch while people tell you it's impossible, it's never going to work. And then within a few years, it's suddenly an industry category that analysts are following, customers are buying into. It's a fantastic feeling. And right now the Viper is getting a lot of traction. So tell the folks out there, where you guys fit, the Scale.io piece, because I just asked a question to Sal. You know, where does Scale.io, VSAN, and Viper, how does it all fit in? People want to kind of know the differences of what you guys fit into where. Okay, so Viper is, Viper Controller is a management and provisioning tool that makes it possible to manage heterogeneous storage environments in the data center. And Viper allows you to manage both the what we call the platform two traditional storage arrays, as well as the software only stored systems like Scale.io, Viper Object, et cetera. And the beauty of Viper Controller is that it enables customers to start by facilitating the operations and the management of their heterogeneous data center today. And then to continue to use that same management platform and to bridge into the third platform type of software defined storage. Now Scale.io is a scale out software only block storage solution that runs alongside applications, databases, or hypervases on the actual application servers. So Scale.io is a server-send software solution. So, let me ask you, so about five years ago, Brian Gallagher invited me to come speak to his, his Woods meeting, you know, his offsite planning meeting. And I said, I just want to talk to you about one thing which we think is going to be a really disruptive trend. Of course, we're talking to the high end group. So we laid out, we said, we made a premise. We said that for 15, 20 years, functions have been moving away from the server out to the sand and for good reason. And now we see it moving back. And we see flashes, this big disruption, and it doesn't have to be flashed, but that's one piece of it. Moving things closer to the server and this distributed set of coherent nodes, somehow interconnected. And I got a lot of heat, you imagine. Five, six, seven years ago, right? How are you going to do this? How are you going to, you know, make it coherent? How are you going to communicate? What's the fabric and all this stuff? Yeah. And I said, I don't know. I don't know how to solve it. I'm a technical guy. Yeah, that's your problem, right? You did that. So take us back to when you started the company. When was that? So we started to work on Scale.io in 2010. Okay. And what we saw at that time was that the traditional way that we are used to build data center architectures with three layers, servers, fabric and storage, started to not make sense anymore. This type of architecture was actually a workaround 20 years ago when the application servers had hardly sufficient resources to run the application that they needed to run. So there was not enough resources in the server to also run some storage management software that would take care of resilience and advanced functionality, et cetera. So get it out of there. Yeah, so what people did, they took other servers, filled them with storage, disks, and ran a dedicated storage software on it, which was the front runner of today's San market, the $50 billion market that we have today. Now, in the last few years, servers have more than enough resources, CPU, memory, bandwidth, local capacity, and you can run multiple applications and databases. Even virtualized. Even virtualized, and still have enough resources to run the software layer there. And so if you can do that, then it doesn't make a lot of sense to spend all the operations and all the capex on an additional layer. So this was our original premise when we started Scale.io. Obviously to do this, you need to take care of a few little things, high availability and resilience, high performance, and all the enterprise class capabilities that people got used to with the enterprise storage systems in your software layer that runs on the application servers. And so this is what we did. And so the other basic concept here is to take the storage that's within the server and pull it. Correct, yes. And then so how do you communicate? That was the big question I got five years ago. How are you going to communicate across those nodes? And how are you going to move things fast enough so there can be coherent? Where's the single point of control? So how do you do that? So the architecture of Scale.io is a distributed parallel architecture where what we do is we stripe the data over a large number of devices and a large number of nodes. And then we aggregate both the capacity and the performance of the various nodes and we present it back as a virtual send to the applications in the environment. Okay, so when you think about the markets for this and we sort of laid this out, we have actually a piece on Wikibon. It's just if you Google server send and Wikibon, you'll see this, but we laid out sort of the marketplace. You've got traditional arrays, you got cloud storage, you got Das, flash only arrays, PCI server flash, you got hyper scale. And so we asked the question, where over time will server send fit? And there's not a lot of white space where you don't fit. I wonder if you could talk about that. Was that your vision that you can, essentially your total available market is the entire storage market? Yes, this was absolutely our vision and we are executing on that vision today. So the way that we build Scale.io, Scale.io is the most scalable solution in the market today so we can scale from three to thousands of servers in a fully elastic manner and we do not force the customer into any type of hardware lock-in or into any type of configuration lock-in. So the customer can run Scale.io on any server hardware and server configuration in terms of CPU, memory, network bandwidth, type of disk, size of disk, number of disks in the server, et cetera. And we also don't force the customer even into any symmetric node. So he can have different types of servers in his data center and still aggregate all of that with Scale.io into a single storage pool. And by doing this, we actually enable customers to build, for example, magnetic disk-based, very large capacity environments or very high performance PCIe-based, flash-based, low latency type of environments and everything in between. Right, so cheap and deep or active data. What do you make of open source and some of the movements in open source? Things like Ceph, for example, you see Red Hat just got it through the game. What's your take on that? So first of all, open source is great, it's important. We are not, we didn't go down the road right of open source because we believe that stored systems must be robust, commercially supported. You say you didn't go down the road of open source in terms of making your product open source. Correct. Do you use open source in your development at all? No, we build everything from scratch. Okay, okay, so. And so Ceph is mainly an object-based solution as we see it. And it's probably a very good solution. Okay, so your approach is to get the market faster, more robust, better service, that whole model. Where do you see initial traction? We see traction in a few environments. First of all, we see a lot of traction with both private cloud environments in the enterprise and public cloud environments with service providers. We see a lot of traction with customers wanting converged environments, whether it's OLTP database is running and converged architectures or hypervisors and VDI that people want to run in converged environments, sometimes even in appliance type of approaches. And so we see a lot of traction in all these type of directions. So we say converged. You're talking about convergence of compute storage and networking? Yes, absolutely. Talk a little bit more about how you play in that converged infrastructure market. So Scale.io was built to be a very small footprint solution so that it can run on the actual application servers alongside the applications, the databases or the hypervisors. And by doing that, and by enabling the aggregation of local capacity inside the servers, we effectively enable customers to create hyperconvergence or to enable also partners to create hyperconverged appliances that they are selling in the market space. So we've talked a lot on theCUBE about enterprises trying to replicate hyper scale. Is that a trend that you guys see happening and what does that mean, bringing hyper scale into the enterprise? So we see that mostly in cloud type of environments, whether it's private clouds or public cloud. So people want to create platforms that are generic platforms enabling their users to get as many resources, compute and storage resources as they need in a very easier manner. And so when you go down that route, you want to just be able to add more and more servers that represent storage, compute and network without having to take into consideration the traditional challenges of capacity planning, performance planning, et cetera. You want this to be just very elastic and simple. This is what scale enables customers to do. So what's it like being part of the 60,000 person company? A big culture shock or you still get your own little development team in Israel and fairly autonomous or what's it like? We try to combine the best of both worlds. So we are putting a lot of efforts on remaining fast and agile with our development and with our sales and marketing approach from the scale perspective. But on the other end, we're doing a lot of efforts to integrate with EMC products, Viper, RecoverPoint and others and to integrate obviously with the EMC organization. So our integration into EMC and in terms of organization, finance, HR, et cetera was one of the fastest in EMC's history. So give us a quick update on what's happening in Israel. It's a hotbed of innovation has been for a while. It's kind of the new emerging mini Silicon Valley. Why is that and give us the update. I don't know why that is. Lots more people, a lot of well-educated people, I guess. But there's something else there, right? Well, I think there is a combination of good engineers, financing and a lot of good people that are entrepreneurial and that want to take their ideas and execute on them. Maybe we're the military together. Yes, you get a lot of that as well, absolutely, yes. Excellent. Boas, thanks for coming on theCUBE. Really appreciate it. I want to ask you to share your thoughts to the folks out there on, you're giving as an entrepreneur, you've been an entrepreneur now at the big company. With all this change, why is this point in time in our technology history so important? So I think we are experiencing a big shift in approach to data centers, which is driven by a few things. It's driven by the understanding that actually, the understanding in enterprise environments that actually much larger data centers that needs much more SLAs and much more performance in the web 2.0 are actually able to do that on completely commodity hardware. This realization makes IT executives think twice about how they build their data centers today. And I think there are other drivers related to the belief that convergence translates into easier operations and better manageability, combined with a lot of pressure on doing more with less. And so I think this combination of drivers in the market is pushing this whole move from the traditional three layer architecture to a new architecture that is scale out, commodity based, high performance and a single layer. Thank you so much, great insight. Great to get the tech athletes like yourself coming in and I know you worked hard. Congratulations on all your success. Great to have you on theCUBE. Appreciate your thoughts. This is theCUBE, we're live in Las Vegas for EMC World 2014. This is theCUBE. We'll be right back with our next guest after this short break. Thank you.