 Live from Boston, Massachusetts. Extracting the signal from the noise. It's theCUBE, covering HP Big Data Conference 2015. Brought to you by HP Software. Now your hosts, John Furrier and Dave Vellante. Hello everyone, welcome to Boston, Massachusetts. This is theCUBE's special broadcast here at HP's Big Data 2015 conference. This is our third year here. This is theCUBE. I'm John Furrier with Dave Vellante. My co-host for the next two days, wall-to-wall coverage. Again, live in Boston, Massachusetts. This is exploding into a full-fledged HP software event. Three years ago, Dave, when we were here, this was a skunkworks, a guerrilla operation, part of the Vertica Big Data operation. They had huge success with that acquisition. Vertica was one of the best successful acquisitions in HP history from a price-to-value perspective. Now it's exploded into a full-on conference. And really is, the combination of DevOps meets Big Data, meets developers, meets customers. These are the cutting-edge guys working on some of the hardest problems, from security to entertainment to commerce. Big Data is now everywhere. Dave, great to see you again, and great to be in Boston. John, it's been awesome having you here in Boston. You spend the time in the Wikibon offices. I really love to have you and shaking things up. And it was great to sort of see you last week and this week here at the HP Big Data conference. You know, when HP bought Vertica, I was struck by the lack of attention that the current then management, you know, Leo Apatek at the time, paid to HP. I mean, it was just this diamond in the rough and they didn't know what to do with it. Finally, it took a couple of years, but HP has really started to see the potential of HP Big Data and specifically Vertica. And I think Meg Whitman has put a lot of effort behind Vertica. Now, here's the thing. HP has taken this collection of assets, Vertica, Autonomy, their activities in Hadoop, their enterprise security, and this sort of fuzzy development platform and put together Haven. And I think what users are looking for here at this conference is how is HP going to help them build out their data analytics pipeline? How is HP taking those sort of disparate pieces and putting them together for analytic solutions? Colin Mahoney this morning in the keynote talked about composite analytics, basically taking the piece parts and making them more consistent and making them more of a solution. He drew an analogy, John, with ERP. How ERP went from highly customized to more packaged. And I think we are, in fact, George Gilbert has said we are following a similar track with analytics. However, HP has a slightly different take on that where the package has to be more flexible. It has to be more malleable. So I think Colin Mahoney is talking about a different type of scenario that we're going to talk to him about. And of course we saw in the keynotes this morning, John, a lot of great content. What were your thoughts? Well, first of all, the keynotes were great. You had Robert Young-Johns kind of out there kind of with a swagger. I couldn't, I didn't know if he was clever or kind of being goofy on purpose, but he kind of hit a nerve there. He kind of was the non-HP executive telling jokes. You know, he made a comment about VCs, trashing the hype about big data. Brought kind of a practitioner vibe for EVP. I don't know if he's just fast and loose because HP is re-organizing, it's all new fresh blood here, the new organization. But I was impressed by this conference's keynotes. Colin Mahoney then came on, delivered a very strong speech and keynote. Obviously he was the CEO of Vertica, which was, again, one of the best acquisitions in HP history. You look at the value that they got out of Vertica, certainly compared to autonomy night and day in terms of like cost of value. But then they interviewed Mike Stonebreaker. It's the Mike Stonebreaker Cube alum been on a couple weeks to go here with theCUBE. Again, really, really raw, awesome content. And they were letting it all hang loose. And really the data was great. And they talked about the trends, they talked about volume, velocity, variety, and really broaden the whole scope of big data. But what I'm most impressed with Dave is this conference is about developers. And it has that, it's not headlining as a developer conference, it's headlining as a customer conference. But Vertica has attracted the kind of audience and customer base that are the heavy hitters. You look at Vertica, what Vertica has done Dave. They brought in heavy duty DevOps guys. Facebook, Wayfair, the people we're going to talk to here. They're bringing in large scale. And this is a large scale problem, both on the cloud side, DevOps side, now on the software agile side. You're looking at a whole new explosion of software development, lifecycle changes going on. You're seeing a new way to develop. You're seeing a new way to provision software. You're seeing software lead the effort. And Vertica, really the crown jewel, Dave, we've talked about something theCUBE, attracts that kind of customer. It's not like the guy who's been doing mainframe software, client service software. These are the cutting edge guys at the top enterprises are hiring full stack developers, front end, fully asynchronous, really writing, cutting edge software. So to me, I think HP has stumbled upon an opportunity, lighting in a bottle for them to bring in fresh innovative customer base and developers. Well, I think that's smart. I think that emphasis on the developers spot on it, man. Frankly, that's HP in my opinion, John. They're biggest challenges. How to build that ecosystem of technology partners to plug into open source technologies, like streaming, like Kafka, like Spark, and then how to build out some systems integration expertise. And then most importantly, how to attract developers and development. So one of the things I want to talk to Shupa Luanda about is what is that development platform look like inside of Haven? What are developers actually utilizing? Is there a cloud foundry like Paz Layer? What is that layer? She talked about some partnerships there. I want to dig into that with her. The other thing, you mentioned Stonebreaker. He was awesome. He was like throwing BS, cold water on the whole big data theme. He took us through sort of a history of the modern big data where Google came up with MapReduce to try to solve an indexing problem in search and then quickly realized it was going nowhere. So the industry pivoted to HDFS and now the industry's pivoting to Spark and then there's something else coming. And so he really warned practitioners in the audience really not to buy the hype and really pay more attention to what, I think Young Johns was talking about, what is your big data strategy? The technology will come and go, but what's the strategy to build out your data pipeline? And then again, the big question that we have here and we're going to explore is, what's HP's role in building out that data pipeline? Couple observations in the keynote for the folks that weren't in the audience. Couple jabs going on. I mentioned the VC jab about the hype. Really trying to bring it down to a grounded level. Both Robert Young Johns and Mike Stonebreaker both mentioned hype. The other thing that was interesting is they mentioned Tableau twice in kind of a jabby way. I mean, we're going to go look at the tape, but I think there was kind of a subtle Tableau depositioning in the keynote. But it was more than subtle. I mean, I think basically Young Johns said, you know what? You can build pretty graphs and charts and make them spin and do all this cool stuff, but what's your strategy? What are you actually trying to get done? It's not just pretty viz. It's all about- Do you find that ironic that Tableau is not attending the conference and then two references in the keynote to Tableau? Well, I think that's more than ironic. I think there's maybe cause and effect there. I don't know. I mean, you know, maybe Tableau sort of eking into the vertical space. You know, Tableau is basically essentially just a visualization package. Of course, the company's doing great, but what else is there? Are they building out the stack? And so is Tableau in vertical? I'm a big fan of Tableau, as you know, Dave. I love Tableau, love the company, but I just see a cloud over Tableau right now, almost as if Tableau is stuck in the middle between two choices. Classic competitive strategy stuck in the middle. Do I go build a full-on platform, or do I stay to my roots and my knitting, which is visualization? And if you believe young John's kind of, you know, jabs at Tableau, you can say, hey, you can't be both, either pick a position. You've got to go, you can't be half pregnant in the market. Well, but with a seven to $8 billion valuation, Kristen Chavo, the CEO of Tableau, has to be thinking about expanding the tan. Yeah, again, innovators dilemma, right? Do you stay the overlay graphics, or do you actually innovate on top of it? Now, you start innovating, then you're building your own stack out. Here we go, back to our conversation. Is there platform wars going on, Dave? Well, this is perpetual platform wars in this industry, and they're heightened at certain times. I think this is a time more than ever where there are platforms of why? Because the cloud, you know, this whole social explosion, the data analytics pipeline, the big companies getting disrupted. So yeah, absolutely there are platforms. And we will be drilling down on the whole platform wars. Look to wikibon.com, siliconangle.com, of course, theCUBE. We've got VMworld coming up, and OpenStack SV in Silicon Valley, where this is the highlight of the conversation. This is the most talked about thing. What is the cloud platform that will power the next generation analytics? Again, a lot of stuff going on under the covers here in analytics. DevOps and cloud are powering applications. Applications are having data native. Inside the application, you're seeing, Colin say, it's not just about pre-package apps, it's about customization, it's about composite. So question, were you surprised by the Stonebreaker comment about calling MapReduce and SQL on Hadoop kind of like the junk drawer? He made a reference. What is your take on that? I want to get your opinion. And this is what we want to explore with you. We want to, so essentially what Stonebreaker was saying is look, all this hype about MapReduce and then the HDFS was just a file system and now Spark trying to bring this into real time. His point was, it's all about the data warehouse. And I think I would add to that point that the Hadoop version of the junk drawer, the bit bucket, if you will, is 10 cents in the dollar in terms of expense relative to the traditional EDW. And that's what's happened is we're seeing the EDW expenditure shift toward a much lower cost junk drawer. So the question we have for the practitioners here, and there are a lot of them, we have a lot of practitioners coming on theCUBE, a lot of Vertica customers is, do they see that the same way? Is this about the recasting of the data warehouse, the traditionally looking back? And can we get that data warehouse to look forward? Or is it just about cutting cost of the data warehouse? So what's your take on the separation between databases, fast databases, and have we moved yet from the old operational analytics where it's been mainly reporting in BI, business intelligence, from a historical perspective? Well, that's the big problem. I mean, you heard Stonebreaker kind of like making fun of data lakes. Of course, we love that, because I always say data oceans as a way to kind of say, hey, it's a bigger conversation. What's going on, Dave? Have we crossed over and what's the research agenda look like with Wikibon on this notion of, hey, operational analytics, nice, that's good, you know? BI, okay, we've been there, done that. Where's this next generation innovation? Well, I think the history of systems of record and the databases associated with those and the analytics associated with those, largely being embedded and tied to that database and the data warehouse. And I think very clearly you're seeing an attempt from the industry to really drive innovation, push that to real time. And that's the promise of so-called big data, but we're not even close to there yet. I mean, Spark just really sort of hitting the mainstream. And it's interesting, Stonebreaker noted that about 70% of the Spark activity is around SQL. So you're seeing the old world and the new world coming together. The big question I have, John, is will Hadoop and big data and analytics live up to the promise that the data warehouse never did live up to? 360 degrees of the business, anticipatory types of operations, operational analytics. And I think in my opinion, we're clearly not there yet. Well, I'm going to share my takeaway from the keynote and just some kind of what I kind of gleaned out of and want to get your take as well. And the segment is, and then this intro is, what? I mean, I'll get to, I loved all the keynote Dave, but to me, there's a couple of things popped out of that. I kind of saw the little glimmer of data that I'm focusing in on. Stonebreaker made a comment about variety. Variety is going to be the big problem. Variety being multiple data sources. So if you believe what he said and we do, and we believe our own analysis at Wikibon, that okay, data warehouse, data lakes, it's just data warehousing recycled in a modern era. Okay, that's data warehousing. But when you're looking at internet of things, self-driving cars, real-time security, application-driven, data-driven applications, real-time, you're going to have this variety problem, right? So you need low latency and variety. And I think that is the nugget that I took away from this keynote was, there's a bigger picture going on around variety. And that's a huge exploration area. And I think that, you know, Stonebreaker made it sound like it's all just a big, big bit bucket and it's just sort of recycled data warehousing. I don't believe that. I do think that our research clearly shows that organizations are trying to become so-called dinosaur marketing term, data-driven, but they're building out so-called systems of intelligence and they're extending their existing systems of record to try to ingest real-time data. And that's a variety problem. That's what Tamer, Stonebreaker's new company is trying to solve. There are others obviously trying to solve that problem, including the open-source community, by the way. So it's, to me, the locus, the epicenter of the organization is shifting from those systems of record to the so-called systems of intelligence, and that's what data-driven is all about. Systems of intelligence, obviously, the cutting edge research from wikibon.com. Go to wikibon.com for all the latest free research and now a new subscription service. If you want to keep up to date on the real-time stuff, what's happening, every day go to wikibon.com. Of course, go to silkenangle.com and watch all the exciting content there, the changes that are going to come in, a lot more content flowing on our site on silkenangle, and if the silkenangle.tbs where all the Q videos are. So we're here, again, live in Boston, Dave, two days of coverage. What do you hope to expect to hear and get out of some of our guests this week? Well, again, the heavy dose of practitioners, the customers are coming on. I want to hear from them. What are they doing with their existing data warehouses? Where does Vertica fit? How is HP helping them build out the analytic pipeline? What is their big data strategy? How are they transforming their skill sets? What is the key barrier to them succeeding in building these systems of intelligence? That's what I want to explore. I also want to hear from HP executives like Colin Mahoney. What's HP's role and how are they differentiating and competing in the marketplace? And certainly the big news on the HP front here is that HP software now is the total group that handles Vertica, the integration of all their big data assets, all their software assets under HP. I want to hear from them. How are they going to compete? I love Colin Mahoney's presentation. I think they're right down with APIs. And so we're going to try to get that done. So this theCUBE, we'll be right back with our first guest here, live in Boston at this short break.