 Live from Las Vegas, Nevada, it's theCUBE at HP Discover 2014. Brought to you by HP. Hi everybody, we're back. This is Dave Vellante with Jeff Frickin. This is theCUBE. theCUBE is our live mobile studio. We go out to the events. We extract the signal from the noise. Chris Selland is here. He's the head of business development at Vertica. Chris, I love having you on because you're a former analyst, right? So you got that angle going and you know what we analysts like to talk about, how we think, but at the same time, you're at the forefront of innovation inside of Vertica, driving new value for customers. So first of all, welcome to theCUBE. It's great to be here again. And yeah, it's fun having said on your side of the table and then being on my side of the table as well. I guess here it's just one table. Well, it was a lot of fun last year at the Vertica user conference. You had a bunch of, you know, myself and Vinny and the database guru come on, we'll help me out. Oh, Kurt, Kurt Monash. Kurt Monash, up on stage, that was really fun. So I want to thank you for inviting me to do that. And I thought it was a really good segment. You moderated it. And so you've got that angle. So I want to start with, put on your analyst hat for a minute. What's happening in the marketplace? You know, we just did this survey, Jeff Kelly's, you know, Wikibon big data survey. An enormous percentage of people said they're shifting resources from traditional EDW into Hadoop. Like 60% said they've done it already. Another 30% said we're going to be doing it by the end of the year. That is a massive, the only one, only 1% said big data is a buzzword of unclear meaning. We're just seeing a massive shift toward the market that you're right in the center. And so what's happening? Well, first of all, I think it's clearly signaling that big data is real. It's very real and organizations, I always say companies, but it's also true public sector, non-commercial. There's so much going on in government, but across the board, everybody's realizing that there's a very fundamental shift going on today in how business gets done and how government gets done and how things get done. And it's all around analytics and it's all around understanding our customers, our suppliers, our operations. And you know, there's just more and more and more data and how good I am at analyzing that and putting it together and putting it in a forum where the leaders of my organization can understand and use it. It is really the key. I think what was interesting, I actually was looking at the survey that you guys just completed. You know, you do have a very high percentage of organizations claiming they're doing big data, which is great. And yes, we were together at Hadoop Summit last week. At the same time, you look at who's using what and from a Hadoop standpoint, yes, there's a lot of exploratory work going on and how can I use some of these new technologies like Hadoop. But yet you sort of say, well, who's using like a commercial distribution as opposed to a free one and very large percentage are still using a free one. So nobody's paying for it. Right, nobody's paying for it. So if this was true. Well, about 25% I think. Yeah, I mean, there's absolutely some. But it's not the majority. It is not the majority. No, it's far from the majority. And what it's really indicative of is that, and this is what we hear all the time from customers, is that they really haven't figured out exactly what they're going to do with all this technology yet. There's a desire to upgrade what they have been doing because they're starting to realize the infrastructures they have in place, you know, there's sort of legacy enterprise data warehouses are running out of gas. They can't support these new data types. You know, all of this social data, machine data, Internet of Things data, you know, traditional EDWs were not built to handle any of this and they can't scale and they don't perform and they're way too expensive. So they want to get on the new technology but they haven't really figured out how. Yeah. So Jeff, that's probably, that's the interesting thing about Vertica is you, there's the other bridge right between kind of the old and new. You probably saw this at Hadoop Summit. Yeah, I was going to say the other two factors that came up a lot at Hadoop Summit is one is changing the temporal aspect of the analysis of the data from getting the data, then reviewing it, then making a business decision to actually collecting, viewing, analyzing the data in real time and making a decision as things are happening, completely different shift. And I think that's what unlocks a giant value creation in the technology. And then the other piece is moving it down the people stack down from the data scientists, down from the PhDs into the line of business folks who can understand it and make intelligent decisions that aren't necessarily earth shattering or sweeping or giant strategic shifts but getting their job done better on a day-to-day basis. And again, I think another factor that just unlocks the huge value that the big data promise really offers. So that requires ecosystem, it requires partnership. So how do you, as Jeff's saying, how do you get the data? Because right now it's confined not right now but a year ago or so is confined inside the data science and the data geeks. How do you get it into the hands of business people that can actually act on it? Right, well to sort of pick up first on what Jeff was just saying and what you just picked up on as well. You know, the terminology we've been using at Vertica is three words, store, explore, and serve. And you know, I think right now, first of all, you can't do this unless, A, you're keeping all of it. I mean, it's, we, and I'm sure Colin, you know, talked about this as well. We talked about Dragonline. I haven't had a chance to see a segment yet. But you know, we hate when people throw away data. I mean, there's no good reason to do that anymore and that's where technology like Hadoop is so powerful because I can have very low cost, you know, basically keep everything. So, but storing it and just sort of putting it in one place, particularly if you don't have it in a form that can be analyzed isn't enough. So then, then you got to be able to explore it and look at it and see what I got and figure out what I might be able to do with it. But then ultimately you don't get value unless you serve it and you have to serve it up to not just, like you said, the data science display, huge role in everything we just described, but ultimately it's not creating any value for it unless you can serve it up in a way it can actually be used by the organization. And again, I have this tendency to always say, used by the business, but there's so many commercial, nonprofit, public sector. I mean, the work we do with people like Conservation International, Wildlife Data, there's a myriad of opportunity, but if you don't ultimately serve it to the leadership, to the decision makers you can actually say, okay, how do I change my business or change the function of my organization? You're not getting any value out of it. And then, yes, to your other point, yeah, there's a whole ecosystem forming here and that's where I'm spending all of my time today. I mean, there's so much dynamism in the Hadoop world. It was pretty fascinating last week to see entire booths at Hadoop Summit set up just to recruit people because there's so much stuff going on there, right? So, and, Do us, talent. Yeah, exactly. There were, we weren't even trying to sell anything. They're just trying to set up booths at this event just to recruit. Come work for me. Yeah. And I mean, the Bay Area is obviously a microcosm or maybe, well, epicenter's a bad word right in the Bay Area, but. That's okay, we're used to it. That's right, that's right. He's a Massachusetts guy, you're the Bay Area guy, right? But there's so much going on. I mean, I think it's maybe a little extreme, but there's so much dynamic work going on there and we've spent now so much time with Vertica, making Vertica interoperable with Hadoop, really providing kind of mixed workload management capabilities, so, because as you said, you know, there's the stuff you need in real time. There's the stuff you need at extreme scale. There's that data you're running your business on and that's the business we're like, put it in Vertica, but at the same time, we've now been able to use and we're doing so much work with SQL on Hadoop, cost-optimized storage is a big part of our drag line release where, you know, if you want to keep it in Hadoop and maybe it's data that you need to keep because we're telling you to keep everything, but you don't necessarily need the performance or you have it, you need to explore it, so, you know, you sort of go across that continuum of store, explore, serve and you go backwards and say, what do we need in real time? What do I not necessarily need in real time? What do I just need to keep around? You know, what's the stuff in the middle? So there's different tiers, there's different levels of data, there's a price performance continuum in this industry that, you know, you've got these technologies can help you basically manage your data at a very low cost and it was a great example of that and then you've got these sort of extreme performance tier but there's all the stuff in the middle and so that's what companies need to make decisions on and it all comes down to what's the value of trying to get out of it. I like the word serve because that implies it's all been prepped in the back, right? And I'm getting my plate that's going to help me and, you know, do something with it as opposed to just, you know, getting all the ingredients and I got to figure out how to make my meal which is going to be different than your meal and I haven't really heard that kind of description because if it's served up well then that's going to make it easier for me to do something with it. Exactly, exactly. And, you know, another thing that we've also done because as you know, you know, Vertica has been, many of our customers are, you know, they're managing gigantic volumes of data, double triple digit terabytes, petabytes, you know, our petabyte club is growing all the time and but in many cases, most cases I would say as a matter of fact they're managing on behalf of many customers. So let's take a use case for instance that I'm sure is true for all of us, right? We're all customers of let's say a particular cell phone carrier or cable provider or what have you. That carrier, that service provider is managing perhaps petabytes of data. Many of them are. But how much of that data is relevant to Jeff Frick? A very small amount. So, but when you go online and you need information on your bill or something that doesn't look right or you need service, the only data that's important to you that's relevant to you is the data that is very, very small size, right? We've provided a way now that we can basically enhance the performance on that small size of data. We've got something called Project Maverick, I should say, which enables us to do that tremendously fast but at the same time allow the service provider to manage all this data. Same is true in things like ad tech where we've got service providers who are serving multiple corporate clients. And again, what client A needs and client B needs and we need to weigh the segment and serve that business up or serve that data in a way that it's consumable to you. It helps you and answers your questions. But at the same time, still manage the mass of it as well. So, I got a question from the crowd. Please answer it. So we're doing a crowd chat, by the way. If you've got a crowd chat.net. I love crowd chats. Yeah, crowd chat.net slash HP Discover. Somebody says, how big is the petabyte club? I was talking, I was asking that question. Paying full of customers over a dozen? Double digits. Double digits. That's as specific as I'm getting. So, and that's a much different answer than I gave last time that question was answered about it. Asked about a year ago. Yeah, so it's growing. Yeah, and it's definitely a different place in double digits. So, I want to unpack some of the. And your survey, by the way, talks about the same thing. I think it was something under 1% of the respondents to your survey said they were managing more than a petabyte. I forget what exactly. The fat middle was. You surveyed on the same thing. So, I'd say if you want specific numbers, the Wikibon survey's a good place to look. But it sounds like your, well, I don't know what the percentage is of your base, but it might map. It sounds like it's pretty substantial. Well, our base does tend to be a lot of them on high end because we're very, very good at that. So, it's probably a high, I'm sure it's a, as a matter of fact, I know for a fact, it's a significantly higher percentage than responded to your survey. But having said that, you know, it's still, that's a lot of data. So, yeah. A couple of years from now, we won't say that. Practitioner of the Cube said to me one time, how do you back up a petabyte? You know, so. Well, new modes of protecting. You have to. You have to. You have to protect it. You don't back it up. But again, this is where some of the work we're doing to really integrate Vertica so closely with Hadoop is really offering us new opportunities where we can do that. So, yeah. So, because you have to. I mean, it's critical data. Talk about, let's talk about some of these partnerships a little bit deeper. So, maybe some of the ones that you're more excited about, I know you got to be sort of Switzerland, but talk about the nature of those partnerships. You know, if you can name names great, I mean, I know you can, but if you want to, great. Of course, yeah. No, I mean, happy to name names. You know, one year ago here at Discover, we launched Haven. And the whole idea of Haven was, the agent Haven was for Hadoop. So, certainly, we're talking a lot about Hadoop. Our Hadoop partners are usually important to us. And the short answer on who we partner with is everybody. We have done, I've got a session today coming up with MapR, actually. That's the next thing I've got after this, or I think one session later, about midday. Because we've done some really nice work to integrate with MapR, to provide true enterprise class SQL on Hadoop, and leveraged them for some of the capabilities we were just talking about. But we're doing the same work with Hortonworks. We're doing the same work with Cloud Era. You know, it's interesting, because even a year ago when we launched Haven, I would say that if you looked at that point, Intel had their own distribution, which they no longer have. Intel, Cloud Era, Hortonworks, MapR, they looked kind of similar. They're directionally, and their strategies are starting to evolve, and now they're kind of going in slightly different directions. Some of them actually don't say Hadoop that much anymore, even though that's kind of what's the core of their technology. But we are working very hard to partner with it all, because the one thing, and it's not about being Switzerland, and trying to be, but we are trying to be open, because as some customers are saying, every single customer I've talked to about this, and I'm sure Colin will tell you the same thing, when we've said, should HP have its own Hadoop distribution? Even though, as we always like to say, you can't spell Hadoop without HP, but should we have our own Hadoop distribution? No, customers don't want that, because it's so early on, because there's so much experimentation going on, because so many of them, as your survey reveals, are still using the free distros. So, you know, there's so much ahead of us, so we've really decided to, the neutrality is about being customer friendly, because we do the same thing in BI and visualization. We do the same thing in the ETL data transformation world. We're doing the same thing, although obviously we're working very closely with the Helion team, but we're working with multiple cloud providers. We've got a great relationship with Amazon, we've got other great relationships there, we've got a lot of customers on AWS. So, you know, we, yes, we're trying to be Switzerland, but being Switzerland is about being customer friendly, but it is why the ecosystem is so, so important, because what everybody's also realizing on big data is, as much as we, like anybody, would like to say, Vertica is the big data in a box solution, and for some use cases it is, but there's so much dynamism, there's so much technology out in the market, and there's so much experimentation and learning going on right now that we really have to be able to support that ecosystem, and so that's what I do with myself. Well, it's the right call working with those guys. A lot of innovation going on, particularly with the three big distro companies, and I think John Furrier joked at the height of the Hadoop distro, you know, when Disco had one, Intel had one, Fujitsu got one, but Furrier joked that SiliconANGLE was starting Hadoop distributions. We had one, but it wasn't even good. We could, and you know, we'll be besties if you do that, but seriously, I think it's actually a big competitive advantage for HP as a company to have this open approach, and be customer-friendly, and offer freedom of choice. So, and it creates more work on our end. I mean, it's harder for us in some respects, but from a customer's perspective, it's the better route, and so that's the route we've chosen to take and we're really happy with how it's going. So. All right, Chris, we got to leave it there. We're getting the high sign, but thanks very much for coming to theCUBE. We will see you in August, right? It's always a pleasure. Well, you'll see me before August, but I'm sure we'll be back in the future. theCUBE is going to be, and we'll have some of our partners on as well, so it's a big data conference. It's a great show, I love it, because it's in Boston, not the flight of Vegas. Yep, conference.vertica.com. We would love to have you there. One more time, and we're definitely going to have you guys. Awesome to see you. Okay. Keep it right there, everybody. We'll be right back with our next guest. This is theCUBE. We're live from HB Discover in Las Vegas. We'll be right back. Thanks.