 Okay, we're back here live in Silicon Valley. We're at the San Jose Convention Center. This is Silicon Angle and Wikibon's exclusive coverage of Hadoop Summit. This is theCUBE, our flagship program. We go out to the events and extract the signal from the noise. This is day one wrap up here inside theCUBE. I'm John Furrier, joined with Dave Vellante and Jeff Kelly. Dave Vellante, the co-founder of Wikibon. Jeff Kelly, big data analyst. Jeff, this is your wheelhouse, Dave. Another great day co-hosting theCUBE. Guys did an amazing job. So this is a summary in day one. A lot of action. We had the top brass on theCUBE here from Hortonworks. We had Amarawa Dahl on Cloudera. We had some startups. We had just some tech gurus. This was truly a tech athlete performance from the guests. And tomorrow, same thing. We're going to have Jonathan Gray, who's now the CEO of Continuity, Ben Werther, CEO of Platform Upcoming Startups, and a slew of other great guests. So I got to ask you, Dave. First, you know, it's great to come and do these shows because we're out in the trenches. We're out with the guys making it happen, building the industry. We're at IBM Edge. We're at HP Discover. We're at Velocity, Fluent, all these other events. This event here, compare and contrast. This event relative to what's happening out in those other events. And then the top story in your mind here today. Well, John, I mean, we're clearly seeing the maturation of the Hadoop when we first started covering this space. It was what is Hadoop? And then, so that evolved into how can I use it? And that evolved into, okay, who in my organization is really going to own this? And what's my data architecture going to look like? Now, we certainly heard from the community here today that, and Merv, we had Merv on, a substantial portion of folks said they're not planning any kind of big data projects and initiatives this year or the next two years. But that means about 30%, but 70% are. So you have a lot of activity going on there. I think there's still tons of uncertainty as to where that initiative really belongs. And I think there's a dissonance between the business line and the IT people. And the other thing that we're hearing that's a sign of maturation is security, security, security. You put your data into a Hadoop system and all the security policies and edicts go away. And you're essentially starting from ground zero. And so you're seeing the community rush to really provide that capability. We're going to be talking to some folks tomorrow, particularly the guys at Squirrel about what all that means. And then the other thing I think my big takeaway from today is a lot of ways to skin a cat, John. You got Hortonworks, which is the anchor sponsor of this event and the primary contributor here of this event saying, the Lightning Run, it's got to be all pure open source. And then you've got Cloudera, which is kind of in between saying, look, we love open source. We're going to add value on top of that. And we're going to do balance. MapR is even a little more to the right of that spectrum saying, hey, we love open source too, but we're going to drive harder and harder to solve those customer problems. And they all seem to be doing well in the marketplace. As Armour said, in the early days, there were probably dozens of Linux distributions. Today, there's still five or six. And I think he's implying that you're going to see some similar things in the Duke world. Jeff Kelly, what's your thoughts obviously you're covering the space and you had a lot of prep coming in. You're talking to all the companies and their competitors, each other. And you're talking to customers, you're talking to end users. What's your take given that you're tracking this sector? What's your take of Hadoop Summit today? What's your take on what's happening? Well, I think Dave hit the nail on the head with this is all about Hadoop growing up, becoming truly an enterprise-grade platform. We heard that here last year as well, but there's been huge progress since then. We're hearing a lot around data governance, around access controls as day-state security. So these are clearly questions you start asking inside an enterprise when you're moving from a little test environment with a few Hadoop nodes and you're doing some testing and you're experimenting with some algorithms and seeing if you can find some interesting insights to when you start moving to production and you actually want to start exposing real customer data to Hadoop, to these applications you're building on top of Hadoop, exposing those applications to your internal workforce, maybe even outside your organization. And that's when all these questions around security, governance, data quality, et cetera, come into play. You've got to take those into consideration and I think we're starting to see more and more of that conversation happen and we're seeing that here this week. The other big story, I think, is the continuing, what I would say, is a major battle for the hearts and minds of the Hadoop community. Is it going to be open source 100%? Is it going to be some mixed version path or somewhat proprietary with some open source Hadoop components? What is this platform going to look like in a year? Which distribution ultimately is going to win? So that's clearly an area that's still up for debate. It was at the, one of the panel discussions today was around SQL on Hadoop and the different approaches to doing that and got a little feisty there with folks from Pivotal talking to Cloudera, talking to Hortonworks, with, of course, Datamir being the outlier, talking about their approach where you don't need SQL at all on Hadoop, but those are some of the big issues I'm seeing here this week. So John, what are your thoughts on, Jeff, you mentioned Pivotal. You're not hearing a ton of them today. Merv mentioned them as part of the big whales, but what's your take on what's going on here? Well, first of all, let me get the Pivotal in second. I just want to give you my take on the top story here. To me, I wrote on Forbes. I think the main story is that the signal that this ecosystem is strong and vibrant is clearly the $50 million raised from Hortonworks. And there was some speculation from Kurt Monash, a friend of ours, we know him. He's a hardcore analyst. I mean, he was sticking the needle into the market by saying, hey, Hortonworks is going to be acquired, right? It's a lot of scuttle blood around Hortonworks, an acquisition target, Intel, Microsoft, you picked the suspects. They're all been discussed. I think that kind of puts a wet blanket on that rumor that they're going to be bought out. $50 million in fresh financing that they just closed Rob Bearden, obviously aggressive. So that is a signal to the health of the ecosystem. What also is interesting is Amar Awadallah on the queue here, basically saying to the community directly, talking to the camera, and saying, hey, this is about contributing, not just taking, don't just take, contribute. That is an open source ethos that is just going to be driving the marketplace. So you're seeing the strength of the two players, okay? It was once arm wrestling between Hortonworks and Cloudera, both are standing tall and together at the altar here, saying, hey, the dupe ecosystem is strong. Yes, there are competitors and the financing and signal that that's healthy. Now, will they get still acquired? I'm sure there's always discussions about $50 million, Dave, as we said earlier, certainly changes the price of the acquisition depending on when they do that. So that was big. Regarding Pivotal, I think the interesting story about Pivotal here is that they were not here in big force and I think that's a testament to the fact that their presence in the community just isn't there, right? You're seeing some contribution here and there and I think that at the last strata, we saw them make that announcement, essentially creating that tension, Dave, with their comment about the contributors. And so I think Pivotal is just like, okay, guys, we'll still leverage Hadoop. So I think they look at this community like, hey, we'll play, but we don't need to be the rockstar in the community. We'll take Hadoop and we'll leverage it and we'll put it at the Pivotal platform. We talked to them at the Gigi Industrial Cloud event and they're clear, hey, we want to build a solution and an architecture for our customers and the future modern infrastructure and they don't care where it comes from. They're going to do that. So that's my take on that. And I think you're right. I think Pivotal purpose is to say, look, we're EMC, we're VMware. We have venues that are frankly a lot bigger than some of these little shows and we're going to use those and market to those. Look at Intel. Intel is basically the same thing. Intel is a little bit more democratic as they understand how the tech industry works. You know, it is a team sport. More developer friendly, wouldn't you say? Well, listen, my experience in the tech business 30 years is this. It's a team sport. So you know, you're wearing one sweater one day and you could be wearing a Blackhawks jersey the other day. You don't know what team you're going to be on. It's a team sport. People who build platforms are in it for the long, long game. They're in it for the long haul because platform businesses are very lucrative when they play out. And so it's not a short-term arbitrage game. It's a platform, it's a long game building applications. And if that gets achieved, there's going to be serious amount of wealth creation going on. And fundamentally that is the game here. People are realizing, hey, we want to create some wealth. In order to do that, you have to create business value. And I think the value card certainly needs to be played here. It's still kind of like elusive. What's the value? Where's the value? That's why it's so interesting to see the $50 million day to see the coordination in the community and people rallying around Hadoop because it clearly is strategic. So what you take on that? I mean, you know, you just shared with us that it's a sign, it's a signal that markets growing, things are strong. But at the same time, you know, you look at the numbers. I mean, Jeff Kelly, you did the industry's first and actually still only market sizing from a vendor revenue share. Big data, when you look at the pure plays, not a lot of revenue there. You're talking about, you know, hundreds of millions of dollars tops. So was it a case where, you know, these guys needed to do a raise? You know, were things getting tight? Or are they able to fund, you know, they're going to be able to fund their own business going forward. We're still looking for that big, you know, IPO in the big data world. Splunk, yeah, Tableau, kind of. But what about the big platform guys? I think it's going to be a struggle because we're seeing, I mean, the price pressure being put on Hadoop. I mean, it's a race to zero at some point, right? You know, you're looking at Hadoop, HDFS, really the distribution as the storage layer, really the foundation of your big data platform potentially, but, you know, where the real value is going to be driven is up top of the stack in applications that are actually solving business problems. So the question then for companies like Hortonworks and Cladera and others are, so do we start moving up the stack ourselves? Now, clearly Hortonworks is saying, we're not going to do that. We're focused on Hadoop, only Hadoop. We're investing our resources into developing yarn so that we can enable other developers, other engineers to build applications and frameworks on top to take advantage of Hadoop, but we're not going to do it ourselves. That's not our role. Whereas Cladera is developing, admittedly in the open source world, but they're developing their own solutions on top of Hadoop with Impala and with Cladera search. So they're saying, no, we're going to go up the stack. So I think it's going to be interesting how this plays out. I mean, if you look at our market sizing and Hadoop revenue today is a very, very small slice, Hadoop distribution revenue, very small slice of the overall market and it's still a very small slice, although of a bigger overall market in five years from now. So to us, from my point of view, the value is going to come from applications and it's going to come from the services sector when you're talking revenue, vendor revenue. Yeah, so, all right John, we got another day tomorrow. We should kind of talk about a little bit of what we're going to see tomorrow, but you got any other thoughts before we do that? Well, you know, one of our guests we had on here today I thought was an interesting perspective and I was proud to have him on because I've been big fan of Merv at Merv with Gartner. You know, as for all the kidding, I always give him Gartner about, you know, the magic kingdom, magic quadrant. They do good work and Merv is a great analyst. He was on theCUBE. He gave the keynote here in the community. He nailed it and I thought there's three areas. We always talk about the white spaces that need to be worked on and he summarized potentially our findings, Dave, as well, Wikibon, but he outlined it and I took notes there. Three areas, security, governance and administration. That's orchestration, administration. Those are the three main areas on the platform. Get that done, then the developer market will really be robust. So I thought that was really killer. And then I just really was excited to hear Amar Awadallah on theCUBE, he looked good. Cloudera feels good right now. You're seeing Cloudera looking good and you know, just doing the right things. Mike Olson is not here, we're trying to get him on, but Amar also talked about the key things that had to do batch interactive data and machine data, batch data, interactive data and machine data. Those are the hallmarks and highlighting the HTFS core. Those are very, very cool. And of course, Brian's always fantastic talking about messaging and processing and the importance of persistent storage. That is fundamental. So that wraps up a really good summary of what we're doing. And Rob Bearden and Arun Murthy as well. Rob basically saying, look, we got to harden this thing called to do and we got to build the data services on top of that. And Arun explained to us why we need yet another resource negotiator. Okay, well that's wrap up of day one, exciting day. It's really about enterprise, Hadoop. This is theCUBE, this is siliconangle.com. Go to youtube.com slash siliconangle. You'll see all the videos. We're broadcasting live all day tomorrow and the lineup's going to be packed again all day tomorrow. We're going to have some great guests. We're going to have the vice president from Hedapt, another SQL, we're going to talk about SQL. Not a big theme in day one. We're going to have Abhi Mehta from Triseta, always a great guest of CUBE alumni. Charles Zaleski, vice president of Cloudera. We're going to have Squirrel on, Nutanix, MapR. We're going to have, like I said, Continuity, Jonathan Gray, and Ben Werther of Plattfora. And then just some end user customers. Stefan Groshup as well from Datamere who's an original Hadoop contributor. Who's skeptical about all this. A little more than skeptical about SQL on Hadoop. He's basically saying Hadoop was made for batch and will always be batched. So he tweeted me today, I'm dying to have that conversation with you. Yeah, and you know what I love about this event Dave? Controversy here is rabid and this is happening. People are debating, they're talking, and this is really what it's all about. I mean, this is not just people just going, you know, there's a lot of vendor marketing stuff. This is just real people building an industry. Hadoop Summit is about these are the actors, these are the entrepreneurs, these are the companies. The tech athletes making it happen. This is theCUBE. We'll be right back tomorrow. Stay tuned. We'd be remiss of course if we didn't thank our sponsors. We want to thank Hortonworks for having us here and of course Teradata, Squirrel, Adapt, H-Streaming. We had to help me out to MapR, Nutanix, Cubol, WAN Disco, Splunk. And fantastic sponsors. We wouldn't be here without them. So one of the big shout out to them. Really appreciate the support. This is really an example of the community effort and big data. Yeah, we love our sponsors. They allow us to be independent. They underwrite our programming. It allows us to do other events. Editorially, there's smaller events that might not have the financing. This is theCUBE. This is our fourth season on theCUBE. We're excited. We'll be at Strada, Hadoop World coming up as well. We're going to be doing those events. All the events will be out there. Big data is here. We'll come back tomorrow. Stay here. We're live all day tomorrow. Hortonworks, the Duke Summit, live day two. Stay with us for now. That's a good night. We'll see you tomorrow.