 And good afternoon everybody. We're live here from Wikibon World Headquarters in Malibur, Massachusetts. I'm Jeff Kelly, Wikibon's lead big data analyst. We're here to talk to with Winn Disco, an interesting company out of Silicon Valley and the UK. We're here with CEO David Richards. Hi Jeff, nice to meet you. Thanks for coming. And we've got Jagain Sundar as well, CTO, VP of Engineering. Welcome guys, thanks for coming on theCUBE here at Wikibon Headquarters. So tell us a little bit about Winn Disco and how you guys kind of came together. Jane Gary, you were previously with Altosoft, so tell us a little bit, Altostore, excuse me, tell us a little bit how you guys came together. So Jeff, Altostore was founded by myself and Constantine Schwachko and we were building this big data storage appliance. It's a very hot space. We got founded maybe in January of last year. We were in constant conversations with various companies that wanted to buy us. We were in advanced negotiations with a VC firm to take an investment money. Then we met David and probably in the first, within the first 15 minutes of our meeting us, David took us to meet with Alad, Chief Scientist of Vandisco, who's invented this very interesting active application technology. So both Constantine and I have looked at various ways to remove the single point of failure problem in the name node over the last few years. We're intimately familiar with the avatar node, the backup node, the secondary name node, the consensus we came to was that none of the solutions are really adequate or do what they're supposed to do. Within an hour of meeting with Alad, we recognized the value of Vandisco's active-active replication. And I knew that we could enter the market with a value-added distribution that was truly the only and first active-active replicated name node solution. We walked away from the VC firms, we walked away from the Fortune 100 firms that were trying to buy us, and we agreed to be bought by Vandisco instead, which I still think is the best decision we made. So David, tell us from your perspective what drew you to gain in his team and maybe give us a little bit of context, kind of where you fit into this big data world and how you see your technology as a good fit. So Vandisco was founded in 2005. It stands for Wide Area Network Distributor Computing. It sounds like we're into Disco's 1970s pop and doing all this sort of stuff, actually we're not. Our heritage is in Distributor Computing and we floated the company in the London Stock Exchange in June. It was a hugely successful IPO. It continues to be a hugely successful IPO. And as part of that, we promised our shareholders that we believed that we could take our unique now-pattern-to-active-active replication technology and apply it to additional new markets. So one of those markets that we'd identified was a very exciting, very fast-growing, very important big data marketplace. And we went into the market and we looked for a unique experience, unique expertise specifically in the Apache Hadoop space. And Dr. Konstantin Shafko is credited with the maintenance and creation of HDF ASIS, the inventor of the name node. He was in that party of six at Yahoo way back when that first invented and conceived the idea of Hadoop. And as Jigain said, when we first met, it was love at first sight to coin a phrase, but we knew immediately that this was the right company for us to acquire. It all happened very quickly. Both sides came together. The teams have been working together now for a number of months and we're excited actually to be announcing new products in February, which is really fast, speed to be going at. And you know, this marketplace, things are changing so quickly. So it was very important for WAN Disco to be able to get products into the market as quickly as we possibly could. So, you know, we've been covering this market for a while. And of course, there's some of the more well known names in this market, the Cloud Eras and the Hortonworks and MapRs of the world. So tell us, tell our community specifically, really, what is the value differentiation here? What do you bring to the big data space that some of the other providers aren't aren't offering up at this point? And it sounds from my perspective, it sounds like really the key is enterprise readiness and talk a little bit about that. So at the very highest level, we can take our secret source, which is this patented, active-active replication algorithm and apply it to Hadoop to make it bulletproof for enterprise deployments. That means specifically that we have something coming out called the non-stop name node that I'll let you again explain in some more detail what exactly what it is, but that will ensure that Hadoop stays up 100% of the time, guaranteed, cross data centers, replication, the things that enterprises need. The second thing that we're going to offer is an S3-enabled Hadoop appliance, that's a software appliance that de-risks your third-party deployments into the Amazon Cloud, where we allow you to take it from the Amazon Cloud to behind your own firewall. So those are the two really big value-added components. I don't really see us competing necessarily with Cloud Eras and Hortonworks. We will have our own one disco distribution that we hope customers will use, but equally, we will work with CDH and Hortonworks in equal measure. So if you've already made a decision to use Cloud Eras or Hortonworks, but you want a unique active-active replication technology, you want bulletproof availability, then you can use ours in conjunction with those distributions. Yeah, thanks, Jeff. Thanks, David. So let me talk about the non-stop name node first. So this is the industry's first and only active-active name node replication solution. By that, I mean, any of the name nodes can service reads and writes. So there's true load balancing. And if you have to do some planned maintenance, if you need to update Linux libraries on one of the name nodes, you bring that name node down, install your patch, your RPM or your .dev, bring it back up, and it'll catch up and be back in service. There's no downtime at all. That alone, I think, brings a lot of value to this equation. We also guarantee that applications that run on top of our nonstop name nodes, such as HBase, will continue to run uninterrupted. Applications that you run on top of HBase will not even know of the existence of multiple name nodes or that one went down or came back. This is truly the only active-active solution out there. All the other ones out there are passive standby, hot standby, cold standby. All of them take time to come back up and start providing service. There's no load balancing with any of these solutions. HBase is very fragile when it's run on top of some of these solutions. So that's a clear advantage. The second big point that David already mentioned, we have Amazon S3 API compatibility built into our product suite. So if you've got your storage in Amazon S3 and you're using Elastic Map reduced to run jobs there, you can move those jobs to your Hadoop provided by us with no change to code at all. It's an operations procedure where they point the application at your in-house Hadoop with S3 compatibility, and you'll be up and running with no programmer work at all. Those are the two significant advantages we bring. So let's take into that second one for a moment. So tell us what you're seeing in terms of desire for that kind of capability. The idea that you're seeing enterprises deploy in Amazon's cloud for a variety of reasons, quick to get up and running in a test testing environment is really fairly affordable. I mean, that's certainly the elastic nature of the cloud as well. So why are you seeing or why do you think there's a market there for kind of bringing those back in-house into private cloud? So there are various considerations, cost, privacy, confidentiality, availability, and different enterprises place different values on each of those. But let's talk about each of those costs. It's very easy to get Amazon storage and compute at 18 cents an hour or 10 cents an hour or 11 cents an hour. But your monthly bill can easily run into hundreds of thousands of dollars if you're doing any serious work. We recently spoke to a potential customer who's got a multi-million dollar Amazon bill, and their desire to get out of that system is very high. There are various enterprises that have very confidential information that they will never put on a network outside of theirs. They still want the development flexibility, the scripting capabilities, the API capabilities, and the modern cloud-like elasticity that Amazon bring, except they want it in their own data center. We have the exact solution for that sort of a problem. There's a lot of work happening to take the value that we see in being able to provide programming ability or the ability to convert a data center into a programmable resource is very high. People will find that more cost effective and more valuable to run in their own data center. So let's kind of shift back to the name node and the idea that you guarantee basically availability for your HPACE and the applications running on top of your HPACE deployment. David, so what translate that into really the business value? What does that allow companies to do now that they have that capability? Well, as Hadoop moves from sentiment analysis batch-based programs to real-time runtime, enterprise runtime programs, it changes the dynamic. And even an outage planned or unplanned, and by the way, there are a hell of a lot of planned outages, as Gianni will tell you from his time at Yahoo. But there's still outages. So actually having complete availability of Hadoop becomes not just a nice to have, it becomes a must have, a mission critical must have. So how much did that outage on New Year's Eve, Christmas Eve, and Netflix cost them, I wonder, probably runs into tens of thousands, hundreds of thousands, if not millions of dollars. So if you multiply that out across all the enterprises in the United States that are currently deploying Hadoop becomes unacceptable. So we think that high availability, and particularly name node high availability, which is a known simple point of failure in Hadoop today, we solve that problem, not just eloquently, but it's the only active, active replication solution over a wide area network to solve that problem. Yeah, I think you're onto something there with the one of the things that I think is holding back and do from kind of really production level, mission critical, supporting mission critical, critical applications, of course, is the ability to guarantee really that those applications are up and running at all times. You know, companies are paying a lot of money to the oracles of the world to do that. You know, if in the idea, I think of big data is to move to a more affordable and flexible way of doing that. And that sounds like exactly the kind of a sweet spot you're trying to in here. Exactly. It's resiliency in the enterprise isn't it's a must have product. So, you know, we're very excited when this product comes out in late February, certainly in Q one, we'll see what the uptakes going to be like. But we're already seeing a lot of early demand for this. Fantastic. All right, guys, well, thank you so much for joining us here at the Cube live from Wikibon World Headquarters in Marlborough, Massachusetts. We appreciate it. When does go, David, again, thank you. Thanks again. We'll see you again, hopefully at the Stratoconference here in our studios or perhaps in our studios on the West Coast. So thanks everybody for watching today. Check out when this go very interesting story and we hope to see you again here soon. Thanks Jeff.