 Live from the San Jose Convention Center, extracting the signal from the noise. It's theCUBE, covering Hadoop Summit 2015. Brought to you by headline sponsor, Hortonworks, and by EMC, Pivotal, IBM, Pentaho, Teradata, Synxort, and by Atunity, WAN Disco. Hey, welcome back everyone. This is our final wrap up of Hadoop Summit. They're breaking down. They're going to have to steal the chairs while we're still live on the air or pull the plug. Whichever one comes first, we've had that done before at Oracle Open World. This is theCUBE final wrap up of Hadoop Summit 2015. Just three days of wall-to-wall coverage. I'm John Furrier, George Gilbert, and Jeff Frick. So guys, what a great show. I mean, it's the coming together of cloud, analytics, all doing great stuff. And also more importantly, we had amazing guests on the show. We had a great social community. We had an amazing set of guests. We had Ariana here. We had the whole team working in concert. And we're super excited to come down on the ground here. And you know what? Just want to just say a shout out to our sponsor. We would not be able to do this at the kind of pace that we do 70 plus events a year. Go out and go where the action is. No story's too small. We will go out there. We will get all the action and share that with you. We love doing it. It's a lot of work. We don't mind putting the effort in, but it wasn't for our sponsors. We would not be able to do that. It would sponsor our efforts who are in our community and it could help us bring more community. It's Hortonworks, EMC. Wendisco, Syncsort, IBM. You name it, they're all there. Teradata, Impetus, Pentaho, they're all there. And I missed one. I can't read that. Pivotal. Pivotal. Kyvus Insights, a new sponsor for theCUBE. We haven't had them before. It's fantastic. If you like theCUBE and you want to help us out, be a sponsor and we love it. And we just want to do more and more. Every dialogue that we make with sponsors is we pour right back into the team and continue to grow and do our best to add more and more value and share that. Whether it's social media, Twitter cards, photos, crowd chats, you name it, we do our best. So thanks to the sponsors. All right, guys, analysis of the wrap up. I'll give you my take and then we'll go around the horn. One, Hadoop is ready for prime time. He's continuing to grow. I think Gartner's survey is under forecasting. I think they say half glass, I feel, but I think that's just playing it right down the line. Mervageon did a great job. I think he's really accurate in what he thinks where the market is, trough of disillusionment, but it's going to be big. I think a lot more than 50% are looking at Hadoop. Hadoop is growing. That being said, there's a world bigger than Hadoop, and that really is a subtext to the show here, that besides Hadoop, there's other things going on. Spark had an undercurrent throughout the show. That's interesting. Also we had cloud. We talked about the cloud. We had a lot of people validating our thesis that the systems of intelligence, this next level of analytics, real time streaming will all be powered by the disruption of the cloud, certainly in the enterprise. So I think this is a great time, and the energy was here. Tons of attendees, more and more customers. So I think it's very healthy, and Hortonworks has done an amazing job of staying really clear on their mission, not wavering standard business model. They're still losing money, but they're clearly going for the bigger, going home. So we'll see how that turns out. George, what's your take and what do you think? So one thing that keeps coming back is really this issue of innovation on one hand and the potential for complexity on the other. And I think ODP, sometimes we get competitors bad-mouthing it. The people within it haven't always done it, haven't always done it justice by describing it, which is really the Apache Software Foundation has done an amazing job in incubating and releasing individual projects. But when you get to a Hadoop product, it's really an ecosystem, not a product. And so we needed something that starts to unify all those individual components. And ODP is a very important first step in that. It doesn't take care of the whole thing. It's just about four core elements right now, whereas a Hadoop distro has somewhere between 17 and 24. But Hadoop, in addition to having the core ODP, they have to start wrapping tools, administrative tools around it so that mere mortals can get it running. Or we see cloud options like Microsoft with HD Insight, which wraps its own tooling around it. Part of the reason I think we see a lot of interest sort of under the radar in Spark is it's a very simple paradigm where it has one engine as opposed to many different engines and many different interfaces. One engine, multiple personalities. Is Spark ready for prime time in your mind? It's getting a lot of hype. That's good, but I mean, a lot of people are saying it's not ready. Not, it's probably a fair to say it hasn't gone through the hardening process that most of Hadoop has gone through for the enterprise, but that means ISVs can start building tools on top of it. And I think we're seeing them do that, especially because they can support both Hadoop components and Spark and we'll see them sort of migrate more and more over time towards Spark. Yeah, it's interesting. We'll be at Spark Summit next week, so we'll go and firsthand bring the latest update, which we're excited, but I think it's an interesting point, George, that you bring up. Everyone loves open source for the speed of innovation. You just cannot compete with an entire engaged community as an individual company. At the other hand, as you've said, it's got to be hard and it's got to be enterprise ready and enterprises are happy to pay money to have support for their products and they don't want too much complexity. I think one of the guys said today, complexity kills, so it's this interesting juxtaposition to bring Hadoop in, but then bring it into a way that's sustainable in the enterprise and make an enterprise grade ready. I want to ask you, John, you've been coming to these for a while. How do you see kind of the transformation of the Hadoop ecosystem in general and this show specifically as we've been here for like the last four years? I think, I mean, Jeff, it's a good question. I think the thing I see most is just the evolution of the maturity. You know, I saw some folks last night that I haven't seen in a while that were early cloud-era employees, early Hortonworks employees, they're moving on to bigger and better things to do in startups that are getting promoted. People are really growing. So you're seeing an industry. We always said that Dave Vellante was here. We always talk about this when we say, this is an industry at birth four years ago, five years ago. And what's happened is because we've been in president creation, we've seen the evolution. We've talked to everyone. We know the nuances. So we've been kind of seeing the grass grow, right? It's been fun to watch. And now more importantly, as it crosses over as an industry, still a lot more work to do with customers. And I think that's the key thing. The most thing that I would take away from here is it has crossed the, as an industry, we can cheerlead for ourselves and say, hey, you know, we did great, but still the game's not over. Customers don't see it that way. I think that's why I use that analogy of the ball's in the water, take it out, put it in the drop area, hit it on the green. And that's, I think that's when, you know, the golf analogies is about getting the customers up and running in a very easy way. When you start seeing IBM, HP, Oracle, these kinds of players with big purchasing power, pricing power, market power, embrace this, you know it's real. So that's great. The next step's going to be competition. That's when it's going to be really exciting. You know, the Hortonworks, Cloudera thing, as much as the bickering's going on between the two, is a legitimate, you know, stair down competition. It's a Mexican standoff. It's chicken on the train tracks. It's who is going to make it. It's going to be really interesting to watch because enterprise is going to want standardization. So someone has to be a software vendor, someone has to be a standard vendor. So I think that's kind of playing out and we'll see how that goes. Yeah, and I think the whole, like you said, the big vendors that are here, IBM and Oracle and HP, you know, embracing this technology for their customers. And as I think you said, it's not just about Hadoop, a lot of action on SQL on Hadoop, you know, embracing the broader ecosystem and really connecting Hadoop to the other existing systems of record, system of engagement. What's our other third system, George? And the systems of intelligence that the adults have. To really get a one plus one makes three effect. Yeah. Last thoughts, George. I would say that. Before they roll us out of here. Unlike the era where we saw Oracle and IBM and ultimately Microsoft built huge businesses on the foundation of SQL databases, I don't think we're going to see necessarily huge businesses built on a pure Hadoop platform. One, because it's open source and basically what people are going to be paying for support relationships. What we're more likely to see are the big applications being built on top of that. I'm not sure we know exactly what those are going to look like, but I would say our visitor from Bloomberg is probably more someone who's going to be the prototype where they can take a Hadoop extended foundation and build a big business on top of that. I think, Jeff, the thing to me, is this is all about wrapping up the show. And I think the thing that I would say is the battle ground for the data will be number one. I think EMC's got a huge opportunity just to get out of the way and store everything. And they have a software play up to the top of the stack, but the DevOps coming in, that EMC is an opportunity to be a nice layer so that data can move and have data portability across the various platforms. So I think that's a good opportunity. Obviously, Gordon works. And I just think in general, this is an opportunity for everyone right now. I think we are going to look at massive change and if you're not talking about real time, fully asynchronous API kind of support, notification economy kind of thing we're talking about, this is where the apps are. 360 degree view of the customer, it's got to be real time. It can't be near real time. So these are the kinds of data challenges in that's cutting edge. On the other end of the spectrum, it's the boring but important data governance security. I mean, data leak and kill a company. So that's boring, but really crucial that the foundation and the set. So you've got to be bleeding edge and stable. So I think everything in between is up for grabs. And then you add all flash and do it because we go to hardware shows too. And you think about the computing power too. Again, under the covers, drive this thing and then cloud distribution, really exciting times. Okay, guys, thanks so much. Great show, thanks. Shout out to the crew here. Leonard, Patrick, Greg, Ariana who's doing the social, all the team back at the ranch. You know, Matt, all the Mick, everybody, Andrew, Dave, Vellante watching from Boston. Good to see you out there, Dave, Tweet, some stuff. We've got a lot of stuff coming up. We've got the next couple of shows. We've got DockerCon, we've got IBM Spark, Spark Summit, DockerCon, Red Hat Summit. Go to silkenangle.tv to find all the action and all the schedule. So again, silkenangle.com, wikibon.com, silkenangle.tv, and of course, go to crowdchat.net to join the conversation. This is a wrap from Hadoop Summit 2015, The Cube. We're out, see you next time.