 Okay, we're back live for the wrap up for day one of Hadoop Summit. This is SiliconANGLE.com's exclusive coverage and continuous coverage. Wall to wall day one. We'll be back all day tomorrow for extensive coverage. My co-host for this segment is Jeff Kelly, Dave Vellantes in Boston. Shout out to Dave covering the Dell Storage Forum out there, Jeff, standing in. Doing a good job today. Thanks, you know. You get to stamina like Dave. I got a lot to live up to with you and Dave, set the standards. So, let's just do our day one wrap. This is where we get to talk about our opinions, our analysis and breakdown, the show coverage for Hadoop Summit here. First of all, my take is very successful. On a scale of one to ten, I'd give it a good eight, eight and a half. I wouldn't give it a ten because they had a lot of guys walking behind us on camera, logistically not a good situation for the queue. But other than that, I thought the agenda was great, sold out, good vibe, great conversations. All the sessions are packed and jammed. Not a lot of lobby con. What I mean by that is, when you go to a conference, you can tell by how good the conference is, by how good the lobby is. If the lobby is good, the conference isn't good because people are talking in the hallway and those are the best conversations. This conference is all about the sessions. They're packed, they've got meat on the bone, it's technical with business case. So one, I think that's great. Two, the target audience here is not all suits. But there are people here talking business in the sense of use case, market opportunity, not a lot of selling. So that's positive. And just some highlights from the day, from my standpoint, looking at the agenda. I thought the interview with Rob Bearden, who's the CEO of Hortonworks, was fantastic. His first time on theCUBE. He did a great job. He's good. I mean, he's a fighter. You can tell. I look right directly in his eyes. You can see the fire in the belly, he knows the space. It's very complimentary of Cloudera and vice versa. So obviously, there is no real war between the two. That's big news here. And Doug Cutting, obviously genius in his own right. But in general, obviously we had MapR, they're storing Amazon, big story of bringing Amazon in. Yeah, absolutely. Amazon needs some simplicity around that. And just in general, there's a huge entrepreneurial undercurrent between infrastructure and application and analytics. And that is the top story. In my opinion, is that there's work going on in both of those theaters. And those theaters are exploding with innovation. And there's so much disruption around this open source community. The big guys can't even get their hands in here. IBM, HP, they're just sitting on the sidelines waiting for the winners. Well, I think you hit on a really good point. It's not a matter of waiting for the infrastructure layer to kind of become insured before we start to see applications and that integration layer. We're hearing all about integration at the show. And if you look behind us, there are vendors from, you name it, data integration, data application development, big data application development, the infrastructure layer storage. So there's innovation happening on pretty much every level of the big data stack. But for me, the real big story was I think we heard from just about every vendor on here was talking about enterprise ready, making the platform enterprise ready. That's the key to this market. When you get to the point where the enterprise has no second thoughts about deploying Hadoop in terms of supporting mission critical applications and workflows, then it just opens up the entire market to all these other players, all the startups we're seeing out here. So really the goal of the distribution vendors is enterprise ready. That's kind of the end goal. And the other thing we're seeing, of course, is the co-operations you mentioned. There really isn't a war going on, like you said. And as you pointed out so nicely that my estimate of $53 billion by 2017 for the big data market, $50 billion, is probably a bit low. And that means there's a lot of white space, there's a lot of opportunity. And there's no reason right now for a lot of these vendors to go cracking heads. There's a lot of opportunity out there. I mean, I think everyone kind of had a gut feeling, Jeff, that it is. And how you size the market is interesting. I think if you look at the impact of big data, certainly it's in the trillions, right? It's in the trillions of trillions. But because it affects every business from the military all the way to retail. And I think we are all seeing classic innovation cycle here in Silicon Valley and beyond, where there's so much wealth creation opportunity is that you don't want to have a situation where, as the old expression goes, your eyes are bigger than your stomach. And there's plenty of beach head, plenty of room for billion dollar businesses. And that's all getting aside. There are a billion dollar businesses in this ecosystem that will flourish out of this new wave of big data. They're in this room. And they're in this room here. It's kind of like half the people in this room are going to be billionaires. Or certainly successful because if the community continues to align with the stars of big data momentum, you're going to see that. The other thing that I would say that I see as a real big trend out of Hadoop Summit is something that we've been reporting on consistently is the notion of moving from batch to real time. And that what real time means is a very interesting conversation because with Hadoop, you have the whole eternity of HBase, HTFS, and MapReduce. But you can get near real time. So the word near real time is good enough to get solutions out of the door. But still not good enough to cross the chasm, because we're going to talk to Jeffrey Moore tomorrow, to the big deal, which is the customers of IBM, the customers of HP, companies doing billion dollars of transactions on their infrastructure, enterprises that have thousands of servers. So that's where the real deal is. So to me, that's where it's going. And the big guys are just waiting. It's like, okay, back up the truck from the Hadoop ecosystem. We're buying. So my big vision and take away from this is that absolutely a marketplace of buyers will be here. They're already kicking the tires. So huge exit opportunity for the right technology and certainly for the right company, IPO and massive growth. Well, I think we have to talk a bit about, while there isn't a war, there are definitely different business models going on here. Hortonworks taking that kind of that red hat model. We talked to Rob Bearden today. I'd love to get your take. You've been in the Valley for years. Do you think that model is really viable? Which one? The Hortonworks model of kind of priming the market, giving away, essentially, the products. And then as they scale, selling them support. I don't know. The jury's chill out. You have to talk to Rob. I feel differently about my initial reaction when talking to Pat Gelsinger, who makes a very compelling argument that the red hat with the Linux is a different mindset than Hadoop distribution, which is a different community, different product approach. And by the way, the product market is evolving very rapidly. Again, both on the infrastructure side and on the software side. So the quote red hat model assumes a massive land grab by one leading distribution and you saw MapR with their distribution deal with AWS certainly changes the game a bit for large scale clouds. So on one hand, I think that that points to a no. On the other hand, in looking and talking to and seeing some of the smirks on his face, and I can see the wheels turning in his brain, that there's other stuff going on that we haven't yet reported on, like what's going on with VMware, like what's going on in traditional enterprises. And because there's a trillion dollars of value being up for grabs, a big pile of money in the middle of the table, that's going to be redistributed. So if you're an incumbent vendor like an IBM or an HP or a big player, you have so much at stake. I mean, look at what Microsoft's got at stake. The trillions of dollars of business that might change hands. When you have that kind of disruptive marketplace, you could lose trillions. So there, I looked at Hortonworks really being that arms deal. So I think their business strategy is sound in the sense that they're going to have an outcome either way. So I don't think at Pat's point, although legit, there's an argument that says, okay, given the nature of business, there's also the scenario where no matter what Hortonworks does, the outcome may still be the same, massive exit, massive acquisition, certainly locking in those distribution deals. And can they play the, quote, chess game and poker at the same time is ultimately what they need to do. And in talking to Rob, I think he's that kind of executive in looking at the team. We talked to them that they have guys who can play chess and poker at the same time. So I think it's a viable strategy. And I think it's a good call. It's a good solid strategy. I think if I was in their shoes, this is something that I would do as well. Yeah, so, you know, we talked a little bit about somebody mentioned today on the queue, kind of the consolidation we're going to see in this market and you just alluded to it, but almost the first way, the kind of a first way of seeing some of the smaller startups, kind of actually a wave of consolidation in that layer first, what did you think of that strategy? Or is that, sorry, that idea? I think consolidation will happen. I think it's just, you know, this is one of those things where it really depends on how wide the market gets, Jeff. Like, for example, my opinion is, is that it depends, it's like musical chairs. When the music stops, and we heard Tosso mentioning about it's a great time right now, build value, because you don't know when the good times will end. Ultimately, that's what I call musical chairs. When the music stops, you're not, and you don't have a seat to sit in, you're going to be out. So that's when you have vendors that have conflicting strategies. And I found that interesting with our conversations with the data stacks, for example, because I look at what data stacks is doing is they're not the bell of the ball like they used to be. Cassandra used to be really the big deal. Now it's like a NASCAR race. They're drafting behind Hadoop and other more hyped up environments. And given some of the scale that they're doing, ultimately the proof points in their solution will allow them to slingshot around everybody. So that's going to be the jockeying position. So from the startup world, if they're all fighting the same visual interface, for example, on the analytics side, it really comes down to the products, right? So this is an area that's not going to be a, you can't just throw a haymaker out there and expect to be successful. This is an area that requires, as Arun pointed out, assembly-like coding, C++-like expertise. So you can't just throw a product out there and it'd be junk and expect it to be good. So I think what you'll see is you'll see startups get funded and be kind of a hobbling along with the product. They'll get consolidated to another team and AccuHire. Or if there's too many people competing for say visualization tools, Tableau's doing great. And all these other environments are ahead. They might consolidate down. So I see consolidation in the pockets where there's too much overlap. Then I would immediately, as Pat Gelser, find a center point for each startup because there's plenty of beachhead out there. Absolutely. All right, so in terms of what we're seeing with the 10 needs, we talked a little bit about that today. And it's interesting, some of our guests said, well, yeah, we're seeing this is more of a developer crowd and others are saying, well, we're seeing more business here than we expected. So it depends on your point of view, I suppose. But what did you take on the attendees here? Is this a tech show or is this a business show? This is a tech show. This show is absolutely about tech. And it's about, if you look at the sessions, you can see that the speakers and how they got these different sessions, it's all about tech. So I think that's the most important. One of the big complaints about Strata was it was too businessy. A lot of biz dev, people trying to jockey and understand, which is good. There's a show, we need that show for that. But this show is about use cases and it's way too much work to do. We heard from Dead Cutting, Dead Cutting's talking about Avro and some other things. And when you have that much work to do on the product side in open source and on the commercial side, you need a show like this and this is definitely all tech. So, yeah, but that being said, we've had some practitioners on who talked about the business use cases, some of the kind of how they're bringing value. And that was kind of interesting to me. We talked to Cloud today, doing some very interesting things. They've been kind of working with Hadoop from the get go. And they talked about, you know, kind of, right now, kind of two separate environments, kind of for Hadoop and then your analytics and they would love to see kind of Hadoop consolidate the analytics layer kind of into Hadoop. You know, we're hearing from more and more enterprises, you know, would like to, the more you can streamline that platform, the better. The less you have to move data around the better. I mean, I think the one thing I learned today on that was, drive towards not piping the data, putting analytics where the data resides. Where my instinct was hearing, by my instinct then what I was hearing was putting analytics, decoupling analytics from HDFS and Hadoop into another analytics engine. And you've seen that with structured databases to roll out the data, that was one. But actually putting it within HDFS was interesting. I thought that would be cool, so. Yeah, well I think that's a long term question that has yet to be answered. You know, just about every database vendor out there now has a Hadoop connector. But moving data between Hadoop and other systems, you know, is that a long term viable strategy? And I'm not sure that it is. It's certainly a short term fix, but in the long term, a more consolidated platform with less data movement is probably where we're going. In terms of what form that actually takes the race to be seen. Well, the thing that I looked at the Avro project is an indicator to me of what's going on. And each catalog is a very promising direction with what Hortonworks is doing. The metadata around looking at data sets rather than looking at all this Horton code within the applications. Because what developers don't want to do is recode in the application. That's, we're hearing that loud and clear. Okay, so day one, any final thoughts before we break here? Well, it's been a great day. Another long day here on theCUBE, but I enjoyed it. Real innovations here on theCUBE. We had our Studio B, our first run of expanding our coverage where we had on demand people checking in, it's what we call Studio B. We're grabbing video on the show and really documenting the community here. Having a lot of fun. We'll be back tomorrow for day two. Stay tuned, we'll be right back.