 here from Midtown Manhattan. The Cube's live coverage of Big Data NYC, a Silicon Angle Wikibon production. Made possible by Hortonworks, we do Hadoop. And when does this go? Hadoop made invincible. And now your co-hosts, John Furrier and Dave Vellante. Okay, we're back here live in New York City, wrapping up day two of the Cube here live at Big Data NYC. There's a lot of announcements happening here. We got Big Data NYC, which is really highlighting Hadoop World Stratoconference, all the announcements, companies launching, and we're here live right in the first floor here at the work, right across the street from the hill that I'm John Furrier, Dave Vellante, and Jeff Kelly is a wrap-up. Dave, I've got to say day two, very excited we had. We heard a lot of themes about the data platform. That's the big theme today. Yesterday we heard the story about HD Insights and Microsoft getting in with Hadoop and the cloud. Today it's all about the data platform, a variety of different announcements. Fodere announced their data hub, which is a data platform that builds on what Hortonworks had previously announced, a Hadoop data platform 2.0. We hear Juan Disco with the really amazing and continuous operation, nonstop Hadoop. That's another great message. A lot of great stories here today. I want to get your take first, Dave, on day two of the conversations and what's going on in the show. Obviously super crowded, not here right now. We had great guests. What's your take, Dave? Well, I think, John, we're finally starting to see the focus shift from the kind of boring infrastructure further up the stack. While I'm not quite ready, John and Jeff Kelly, to predict that 2014 will be the year of the application, I think that we're starting to get close. And it's clear to me that a lot of the bottom layers of the stack are going to be commoditized, and that's great because that's going to enable a lot of innovation to occur up the layers of the stack. And that's going to occur in applications and it's going to occur in specialty areas. There's no question in my mind that if people can solve some enterprise problems who have certainly seen that in financial services, there's money to be made in this market space. But I think it's interesting, you hear a lot of sort of negativity on certain business models or positive on certain business. To me, the services business model is a great business model. It's going to throw off a lot of cash. I like what Hortonworks is doing. And then at the other end of the spectrum, you got small little niche application players like Trasada that people might say, well, the market's not that big but they're solving problems. And I think that the application's business has always been highly fragmented. Look at the apps on your phone. That is the definition of an app today. Highly, highly fragmented and the idea is to tuck in, solve problems and grow from a position of strength. And I think that's how the market is going to play out. Very bifurcated and fragmented right now and it's starting to come together. We also had a crowd chat today, Dave. A little flash crowd chat. And I just recently- Flash chat. We just had a crowd chat. Who did? It's a flash mob here but it really talks about the killer app. I said being email and the internet created that killer app. Everyone talked about email as that killer app on the internet. It seems to me big data obviously creating the killer app of applications and analytics in particular. Analytics is the killer app of big data. Jeff, I want to get your take on that because we heard essentially the maturization of the platform where you start to see things be decoupled, highly cohesive elements around a platform. That's an operating system. That's what we're going with the data operating system here. You're seeing that here. So what's your take on all this? Well, I think you're right. The killer app is analytics for big data but not necessarily or not simply analytics in the sense of making analytics available to business people, business users. Like we're seeing, I think that's an important part. Things we're seeing from companies like Platformer from Datamere and others. While that's critical to kind of give business users the ability to investigate data to come up with some insights to do their job better. To me, the real killer application when it comes to analytics is actually, essentially analytics informed applications that actually execute some type of function. So it's not necessarily a user looking at a dashboard. It's an application automating a decision and actually executing that decision. It might be a recommendation engine that's building on analytics in the background that helps a call center worker make the best decision in real time. So for me, those are the kind of killer applications that we need to see more of for this really to become a really mature market and really hit mainstream adoption because ultimately it's about solving business problems. I think we had, when we were speaking to Ben Haynes, CIO of Box, he said for me, big data is really about solving problems and I think that's, he hit the nail on the head there. In terms of the data platform, it's interesting. I mean, we're seeing, to me, Cloudera's messaging around the enterprise data hub I think is good and I think, but I think it's more of a, essentially not a marketing ploy. It's really just a new way to say what they've been doing for a while and what companies like Hortonworks are also doing the same kind of thing. Guys, I got to ask you about a room where he came on. He's the co-founder of Hortonworks. What did you think of his talk today on theCUBE about Yarn? You talked about MapReduce being just another app. He's a rock star, right? I mean, here's a guy who's basically the early days of a dupe and he said, okay, we're going to take a dupe into now just way beyond just batch. We're going to bring in all kinds of not only different data types, but also different data workloads. So you're seeing Hadoop become the fabric for data. Hadoop's data operating system, Yarn is the fabric. So John, I mean, you've watched this from the early days. I remember when you first told me, hey, Dave, is this company Cloudera? Cloudera, I'm like, who? What's the name of the company? Cloudera, who are they? Dave, just watch this space and look what happened. What's your take on all this? Well, I mean, obviously, I'm not afraid to share my opinion, but you know I like to put the vision out there. And sometimes when I see the dots connected, I have those aha moments. And one of the things for me on this Hadoop world and big data NYC trip is the coming of age of Hadoop. It's fun to watch four years. We've been president of creation of theCUBE and our relationship with the community was here from the beginning. And we're certainly not going to go away as evidence here. But to me, the big thing that I'm seeing happening that no one's connected the dots yet that I've seen in the public is that the connection between the software defined data center and big data is coming very, very quickly. Meaning all the work we've seen at AWS, OpenStack, VMworld, EMCworld, IBM, all the top vendor. If you look at what they're doing and how they're transforming their converged infrastructure offerings under the hood, the engine of innovation, the innovation strategy of those big guys, the software defined data centers, the top conversations, probably one of the most important work areas and work in progress right now in the computer industry, Dave. And one of the things that WANDISCO has right now as the only vendor in the big data world that I've seen that actually has that path from big data into the software defined data center. They're talking about high availability. They talk about it nonstop Hadoop, which continues operations. Now that's just marketing around, good marketing that around what people care about when they want to go scale from production environments that have been kind of test POCs with Hadoop, financial services, oil and gas. These guys are moving quickly from in production to scale. And I think that is going to be a major hurdle that everyone has to come about integrating in at scale into full on operations. That's on-prem, that's hybrid cloud. That's software defined data centers. I think WANDISCO has lighting in a bottle. And I think one of the surprises of the show is WANDISCO. Well, that's critical. I think one of the issues we always get asked when is Hadoop going to go mainstream? It's when CIOs can feel comfortable that it's not going to go down and it can support mission critical applications in production, at scale. And one of the keys there of course is the high availability along with security. So for me, those are two of the key areas. Checkboxes need to be checked before. So what do you mean by mainstream? What does that mean? It means that the majority of workloads will be developed in Hadoop. I mean mainstream because we'll see. Adopted by mainstream companies, adopted by the enterprise. Adopted by mainstream companies beyond simply. So right now we're seeing a lot of adoption around on the Fortune 100, the really big enterprises around the world. I've all got Hadoop at least in the lab and many in production. But when we start to see medium sized organizations when they can deploy Hadoop and it's not a new, exciting, but a little scary technology. It's something they can feel comfortable putting in their data center just as comfortable or maybe not quite as comfortable close to as comfortable of a relational database. When it gets to that point where they don't have to, where they don't, the first thought isn't, oh, is this going to go down? Then we'll start to cross that. And two isn't the key commercial applications that are packaged, right? Well, yeah. So there's, I think there's two things. One is some of the large enterprises are gonna continue to develop their own applications. But even those organizations as well are also gonna look for those packaged applications where they can put in a point solution that solves a pressing problem. If you can, as a vendor, if you can solve a problem that is vexing for the enterprise that's gonna return, get a good ROI, they're gonna make you money or save you money. And you can explain it simply. I heard somewhere, it said if you, who should you invest in if you're an investor in the big data space? If a vendor can't tell you very clearly and in plain English what their app does, you might want to walk away. So when you can articulate that and it's very clear what the business proposition is and the business value, I think that's when we're gonna start some more apps. I gotta give Claudero some props here because they got the whole capital market started. And obviously Hortonworks within Yahoo was doing a lot of the early heavy lifting and then the Hortonworks spin out really started to make this thing explode. And as well, EMC's acquisition of Green Plum in the way in which they positioned against the cloud era, that helped get things going as well. But the reason I'm bringing that up is the comment that you made about going mainstream. When I heard Mike Olson in 2010 describe what is Hadoop, the Hadoop 101 segment that he did in theCUBE, he was describing a radically different way of storing and processing and organizing data and handling new types of workloads by shipping code, not data. Describe the problem that Google had, how they saw the completely radically different thinking. Now fast forward to when was it, John, with those guys who went on CNBC. If I were not following this business, I would have thought he was talking about mainstream business. I couldn't tell the difference between Oracle and what he was saying. So that's maybe a signal that's going mainstream. Enterprise Data Hub, sounds like Enterprise Data Warehouse. You've got to be a little careful with that language. No, that's exactly what they're going on. No, that's what they're going on. I mean, that's clearly cloud era. And they see the platform beyond Hadoop. One of the things that's clear to cloud era. And I'm really glad that they just make their move. Finally, post the position, they've been down this path. So they're kind of caught in the innovators dilemma. Do they stay pure Apache and fight that battle? They're going to just do their thing. And my prediction is cloud era is going to continue to contribute and drive Apache Hadoop, be part of that ecosystem and be leaders there. And two, but they're clearly articulating, no, no, we want to win the modern data warehouse. The Data Hub is basically the modern data warehouse. They want that positioning. They want to be able to roll in to the market. That's their, and I believe their intention. That's not from cloud era. That's my interpretation. Well, help me with their names. Were you? Well, I think it's good. Data Hub is pretty specific. Sounds like Data Warehouse 2.0. I mean, it's just... Data Hub, well, Hub and Spook. I mean, you can have multiple components with a lot of things going on. You like the name. I like the name. I mean, I like the word Hub. It implies center of everything. It implies central. It applies not just central in terms of monolithic. It has, you know, Hub and Spook as an architecture. It's similar in that sense, but it's similar in the sense because it's centralized to an EDW, but what's very different, however, is an EDW is essentially for reporting against data. It's not for building applications. They're going to have some type of transactional quality to that application. That's not what an EDW is for. So this, it is different. The language might get people a little confused, but I give them credit for trying to name the category. I said earlier, it's not a marketing ploy. That's not the word I wanted to use. It's a marketing strategy to name a category to own it. Is it a pivot? It's somewhat of a pivot. We've been saying... It's a tack. We've been saying on the queue... It's not a pivot. It's not in this thing. Well, last night I said pivot. They're tacking to a buoy. We've been saying on the queue... Larry Ellison came back from down seven to one in one America's Cup, but you're like in sailing and it's all about tacking. It's all about getting to the endgame. I think Cloudera still has that same vision that we've talked in the beginning. They want to be a platform. They could have sold out earlier. And sometimes it takes a few tacks, Dick. Dave, you nailed it when, at, I think at the Tableau conference, you're talking to Armour. How is this different from what Oracle wants to do? And we've said for the last two years that ultimately... The Splunk Conference, excuse me. How, at what point, it's clear that the Hadoop players are going to start to infringe more and more on the enterprise data warehouse world. And no, no, it's complementary. Well, clearly, even in his keynote this morning, he said clearly there's many workloads that we can move to Hadoop. We're taking that away from the E.D.W. And no, no, sorry. This morning Mike said that in his keynote. So there's definitely tension here. And I think we're going to see at some point there's going to be an all out war between the E.D.W. and the Hadoop side. I just remember well, people saying, ah, PCs are toys, microprocessors are toys. They basically took over the world. So I mean, I think if they can compete in a little bit away, they're going to compete in a bigger way because of the economics. So anyway, that's my take. I want to thank everyone out there for watching today. I know we had a couple of technical difficulties on and off. We've got everything back on track early on this morning. So I appreciate the flexibility there. But also more importantly, I want to thank the folks that participated in our CrowdChat today. We had MAPR participating. We had the folks from the Apache Foundation and the people in the crowd weighing in. It's CrowdChat.net slash HW2013 is the transcript. We're going to be going back and doing more of those this week. Guys, thanks for your commentary. And we'll be back tomorrow. And let's thank the sponsors as well, of course. Hortonworks and Wendisco. Yeah, Hortonworks and Wendisco and MAPR, but mainly Hortonworks and Wendisco primary sponsors for underwriting our ability to come here for the fourth straight years. We again, theCUBE was present in creation of this big data in the Hadoop ecosystem. We're proud to be part of it. And proud to continue to invest in independent open source content. We thank you for watching and come back for day three tomorrow here at Hadoop World, big data NYC. We'll be right back tomorrow. Stay tuned and thanks for watching.