 The Cube at Hadoop Summit 2014 is brought to you by Anchor Sponsor, Hortonworks. We do Hadoop. And Headline Sponsor, WANDISCO. We make Hadoop invincible. Okay, welcome back everyone live in Silicon Valley, this is Hadoop Summit 2014, I'm John Furrier, the founder of SiliconANGLE.com, here with the leading big data analyst Jeff Kelly with bookiebond.org, and our Nescus is in heat shoe with IBM, big data, vice president, in charge of all the big data activities, great to see you again inside the Cube. Thanks John. I know you had an official title, I shouldn't know it by heart by now, but it's so official, it's so long. Welcome back. A lot of things involved, thank you. I'm actually excited to be here, and happy we're seeing you outside of Vegas. I know we have a lot of fun in Vegas, but the air gets dry in my throat, as you tell. It does, it does. Are you guys have a booth apparently, so you're in the trenches? We are in the trenches. You know, problem with the open source community, tell us what's happened. We're in the Hadoop space, surprise, surprise. You know what's interesting is just, I would say the growth, I mean MIRV did such a great job in the opening keynote this morning about realistically where clients are. I think for sure everyone's experimenting with Hive and HBase. I think in terms of Spark and some of the other areas like Yarn, I mean there's tremendous contribution, but it's still relatively mature relative to kind of where the audience is and where the skill base is. I think Sequel was an interesting discussion about Sequel on top. I think the other thing that's been growing, at least in terms of my discussions, is really about understanding vanity, you know, who's who and what context and what persona, and then the security, security is a hot topic for everybody, so. You know, I want to say it's fun to watch Hadoop grow up, and that was MIRV and I was on the Cube earlier, we're talking with Chef, really breaking down where we are on the evolution and one thing that we've done with you guys at IBM is we've always been to your conferences where you guys are fluent in the word business outcome. I mean you have a lot of big customers of a legacy business, but that's new words for the Hadoop ecosystem because you're seeing a massive acceleration of mainstream adoption for Hadoop as a platform for big data in a variety of diverse use cases. Well, in general the platform's still immature in some areas, security obviously seeing the M&A work. Absolutely. So how do you guys look at that? IBM's looking down from the balcony on the stage of the Hadoop ecosystem with a lot of experience. How are you guys engaging and contributing in that community? You know, I will tell you a few things. I thought one of the things that MIRV said in his session that was interesting was, you know, you're going from POC to production, production to platform. Platform is when you're really thinking about holistically the architecture and how you extend out. You know, that's a mindset that is core to our DNA at IBM. I mean, you can't even begin to leave, let's say, kind of the innovation around how we build the technology, how the team really thinks through, how the client's going to adopt it without thinking about the outcome first. Now in terms of the contributions, I think our contribution right now has just been partly educating much like many of the other contributors, educating more the broader industry about the value and then extending some of the enterprise capabilities. So security, integration governance has been a big portion for us. Things like understanding the metadata piece, I mean, we started to add metadata cataloging capabilities within not only big insights but also in terms of the way you catalog that, enabling security over structured SQL type data stores but also no SQL and Hadoop and really thinking about how to optimize various jobs. I mean, this whole space is so exciting. Is the SQL stuff going to take hold or what? I mean, it's the year of SQL again. Absolutely. So we're really excited. We are actually delivering this month and we have a beta program going on right now around what's called a big SQL, a big SQL 3.0. And we actually have a session tomorrow at 3.30 on it and it's really taking, you know, SQL to a whole new level. We've done some internal testing on it. I will tell you this is probably the most advanced SQL that will be here in the marketplace. You know, the thing that gets me excited about the SQL is we were just talking earlier about some of the acting guys about some of the voodoo science at the high end of the spectrum. There's a lot of, we look at parallel processing, for instance. There's a lot of black art magic that very few people can code. So whether it's data science or high end parallel processing, the computer science is very advanced and not a lot of people can actually do that so the talent gap will never be solved. This is just too hard. Well, you know, that point on the talent gap real quick, the U.S. Labor Department actually did a recent announcement that there's going to be like 170,000 more data jobs in the next two years, meaning incrementally around analytics, data architects, data scientists, data modelers. And this is a huge growth increase and like you said, there aren't enough skilled practitioners that actually know some of the new technology bases and some of the new query languages and some of the declarative languages, especially for like text analytics or machine learning. And so SQL is great on two dimensions. One is the skill set. There's plenty of SQL knowledgeable professionals in the marketplace. But the second reason is actually performance. I mean, if you really look at what clients are also putting on Hadoop, it includes SQL data and a lot of log analysis. And so if you want to start to improve some of that performance and also do it across, you know, I think the term here a lot of people have talked about was data lake or data reservoir, you know, SQL's going to be a huge accelerator to that. Well, you know, it's interesting that they, you know, the government points out, we're going to create this, you know, 170,000 more data related jobs. But really, if you think about it, most knowledge workers are becoming data professionals in a lot of ways. And you've got to understand how to interpret data, how to communicate with data. And that's, you know, one of the, one of the softer problems with the non-technology problems. I think a lot of organizations run into is kind of getting people to buy into this, hey, you've got to start taking into account, you know, real data and real insights. What does IBM do to kind of help with some of those more cultural challenges? Well, we've been talking about kind of three things actually at an enterprise transformation level. So one is about how do you build an analytic driven culture? One of the things that we've been tracking over the last two years is also the increase of what's called a chief data officer. I've actually had the opportunity to meet several. This includes like, you know, various banks, like Bank of America, Centros Bank, VB&T, you know, sort of a range, as well as insurance companies, as well as healthcare companies. Especially those that are in that bridge of managing everything from operations to risk, to really thinking about the strategic importance of data. Who's using it? How's it being used? How often is it being used? Who's doing things that are unexpected? Are we getting the value that we expect out of it? And so that's kind of an emerging class. And one of the things that we started to do was actually start to interview and create a community around the new category of chief data officers. So that's one. The second piece we're also doing is actually working with several universities globally to actually put together a curriculum, both in the business school, as well as in the technical schools, for certifications and advanced sort of masters classes around various data type jobs. And we've been tracking actually the jobs that have been published in the US markets and some European markets. So I would say that's probably been going on, that's pretty big. And then I just got back from Israel, as well as Kuala Lumpur in Malaysia the last two weeks. And there's a huge push also around open data, right? So publicly available data. So it's been really interesting to see kind of the national frameworks that are developing by country and government, as well as even private sector. Talk about the globalization impact. That's a really good point. I want to drill down on that. We tend not to talk about that much, but I was on a panel at VMWare with Mark Andreessen, Stanford University founder of FireEye, David DeWalt, all these guys, I was on the panel, I was actually watching the panel. Joe Tucci was there, I wasn't on a panel. I was actually there. I don't have my A game. First of all, CNBC was moderating, did the horrible job. If I was moderating, they did a good job. But the globalization came in as a big point because two things, one, opportunity for new products, but market expansion for companies. And also the infrastructure's different in each country. So why do you look at that? Is it true, do you agree? Oh yeah, yeah. So culture, so one thing is, coming from Israel recently, one of the things that I found really interesting about the country was a huge focus around startup. So we actually just launched IBM along with several other companies, and they're not necessarily technology companies, they're actually enterprise companies around a startup accelerator in the marketplace because it's a huge outgrowth, really from, I would say, it's an interesting skill set that everyone has to go through in the military, especially there's some advanced training that's done around technology as an example. There's a huge investment flux because of a variety of reasons. There's a nice cultural linkage in terms of appreciation of analytics and data. And so you see advancements in that way. You see other countries also advancing depending on different aspects, language, degree of certain industries being more dominant whether it's telecommunications, retail, energy, healthcare. But you know what's the most amazing thing is the transparency of the use cases. So if one team in one country sees what's possible that another business did in another country, yeah, they could be technically a year behind or two years behind in terms of the questions that they were asking, but in terms of how quickly once they learn it they can implement it, they become very agile very quickly. So, and some of the telco carriers, especially in some of the growth markets are much more advanced because they're thinking in terms of the wireless structure. They're not, you know, they don't have some of the traditional wired lines. So back to some of the industry trends. I want to ask you about EDW and Hadoop. Some say that this compliment or the survey that Jeff developed is showing quite the opposite that Hadoop is becoming a replacement in the mind of, I think that was kind of, I don't know if I'd go that far. Anyway, ask the question and then I'll interject. But in general, what truly is complementary in a disruptive market? So I want to get your perspective on that. Weird does Hadoop fit in that because you guys have a lot of commercial products. But IBM is very committed to open source with the IBM pulse event and impact is obvious. And Apache in general, I mean beyond just Hadoop, but Hadoop is one of the core open source project areas but our commitment to open source is pretty strong. Historically, and even today. So what we're finding is, of those who have implemented Hadoop, about 60% have shifted one workload or another from an EW to Hadoop. So it's not replacing the entire data warehouse, but they are being strategic about which workloads they do. But there's a shift happening. Let me rephrase. There's certainly a shift. There's not a massive shift, but there's some shifting going on. You know what I would call it? I'd actually call it kind of two steps. So the first step is streamline, right? And streamlining because one is a cost is a huge driver. Yeah, I mean cost per terabyte of managing, leveraging Hadoop is significantly lower. So you can't get away from the math. So the math says, yeah, why wouldn't you consider it? The second piece is about platform modernization. The types of data sets that people actually want to start to understand are varied and it's not necessarily completely unstructured yet, but it's more what I consider like semi-structured data and that's why you see a lot of log analysis because it's questions people wanted to ask and discover but you still don't see Hadoop fully in production at the scale that you see. Relational systems fully in production for the same reason, which is about currently maturity, backup, recovery, archive, security aspects, redundancy, scale, all these things and many companies are really excited about the opportunity but it's also about timing and I can't, there's also another business philosophy. You rarely want to let's say break or move or stop something that's actually working well, right? So that's why use the word kind of streamline what you need and then really think about how do you modernize to really capture kind of the new workload. Yeah, I mean as a CIO you really got to be kind of a portfolio manager and manage the transition and really pick the right tool for the right job but you live in the real world. It's not, the vast majority of IBM customers are not starting with a blank slate. That's right. A lot of investment in other technologies. Well and I've actually seen mistakes on both extremes quite honestly. I've seen some clients that have moved almost too much to Hadoop and actually as a result have lost their jobs because they weren't thinking realistically about the implications of what that meant. All the way to the other extreme where some haven't been as proactive as they mean. They need to be a lot more aggressive and really thinking about modernizing that environment or being more aggressive about the skill set on their team. So you never want to be on either end. Right. So but I would say most of the market is doing a fairly good job. I mean my preference is that we have more vendors, more contributors, more technology and more client use. So what's your focus area for your M and H strategy? Can you share with us who you're looking at? No. Okay next question. Of course I knew you wouldn't answer but that's always asked the questions. It's kind of the answer. Well you know what, here's how I think about it just quite honestly. So even if you look at IBM's investment I'll be very frank. We invest a tremendous amount internally on organic development. So on the order of probably about three to four billion just in the information analytics space alone on an annual basis. And that's on our own internal development. And then we probably have spent, I would say over the last five years on average somewhere around three to four billion every year on various acquisitions in the information analytics space. So our strategy is really about, okay how do we enable, and I'm going to use the word outcome again, how do we enable true outcome and solution transformational kind of solutions for our clients. And to do that you've got to have a spectrum of understanding of the industry dimension. You've got to have a spectrum of understanding of the use case, the adoption, and it means it's a combination of capabilities. And we recognize some of it's going to be developed organically and some of it that we want to go. So the three to four billion you mentioned on the analytics side, is that R&D and acquisition money or just R&D internally? R&D internally. And then the acquisition budget. Is it another three to four that we've spent? Yeah. So you're looking at almost, yeah 50-50 and almost probably somewhere in the order of seven to eight billion on an annual basis. Yeah. I mean I think most people are surprised. We rarely talk about it in all those things but you can actually see it very clearly through our investor kind of materials and our acquisitions. We just share it. We're happy to get that data out there. I just put it on the crowd chat. So we are here at Hadoop Summit. So I want to talk a little bit about IBM's approach to Hadoop specifically. So can you talk a little bit about big insights? What kind of traction you're seeing in the market? You know, obviously there's a lot of coverage of kind of the startup community and the horse race between Clutter and Mortenworks. But of course IBM's got their own distribution and you've got the other pieces that you layer on top. We do. Explain a little bit to the people out there what kind of what the approach is and talk about the traction you're getting in the market. So, you know, first of all, this is also an evolution that's been happening. I mean if you think about the Hadoop Summit, it started in what, 2006 timeframe? 2008 timeframe and sort of the evolution. One of the things we've been conscious about is really how does it augment kind of the client's environment and the enterprise? And so a lot of our contribution has been really focused on kind of readying it for the enterprise. So how do you manage the workload? How do you handle kind of some of the more advanced analytics, right? So we've been probably pushing the, what I would consider pushing the boundaries around the advanced analytics. So MDM, big match, meaning how do you take the probabilistic match and capabilities that are historically in relational systems and enable that paralyzed over Hadoop as a way to actually be able to correlate structured entities as well as unstructured entities and have tremendous insight. Another aspect would be things like security. We were probably the first to deliver SQL as well as no SQL security capabilities across the spectrum from identifying, accessing sensitive information, masking it to auditing and log tracking to monitoring and activity to encryption. So the full spectrum is something that we've been thinking about because we want it to be adopted. And in order for it to be adopted, you have to think about the enterprise perspective in order for businesses that are in really, have a high level of accountability, fiduciary duty, quite honestly, to their consumers and their businesses that they support. Talk about some of the trends and big data that you're involved in. You and I were talking before we came on camera about streaming that you and I both are getting very excited by because there's new forms of data. I'll say post 9-11 had a whole requirement about Homeland Security. You're seeing video surveillance. I think we talked about that one time. Facial recognition going into the clouds, just like Watson's being a big part of that. So what do you get jazzed about? What gets you excited? You know what, it's really about reimagining. I mean, I kid you not, it's really about reimagining kind of the entire business differently and the way we work and operate differently. So I'll give you a simple streams example that has just been unbelievable just this last year. And we've been talking about this, but the scale of it is we've been working in a lot of medical hospitals around intensive care unit monitoring, whether that's neonatal, whether it's intensive care units, post major traumatic surgeries, whether it's in general emergency rooms. And what it is, is I look at it as the ability to not only be able to process, like a single patient generates, I mean, I kid you not, a single patient could generate like 100,000 data points per minute, right? That's a lot of data for the medical staff to even process before they even diagnose what's wrong with the patient. By the time they figure it out, it's almost too late, or they've lost sort of that lead time. And if you can process all of that in real time, you can actually prevent something. I mean, that's like giving people time to save lives. I mean, you're almost figuring out a way to stop the clock in that moment to prevent, I mean, it just prevent fatalities, I mean, prevent progression into various cancer forms. I think that's, operating much more in real time is going to be exciting because it's not just about operating in terms of processing and accessing that information, it's like the analytic applications, the interactions, the visuals are going to be much more in real time. Video analysis is huge, image analysis, I mean, these are things that we're really advancing the needle on, text of voice analysis, streaming voice analysis. These are things I'm excited about. We've actually done a few scenarios around video analysis as well, so. What's your take on the Hadoop ecosystem relative to its maturity? Put an inning on it if you can. Third inning, fifth inning, late innings. You know what, I would say definitely, probably, and I would say third inning. I mean, we're still at such an early stage. I don't think there are enough new applications that are written quite honestly. And the thing is, I look at Hadoop as much as not only kind of data management, repository and set of capabilities, I really look at it as an analytic platform. So what you're really excited is about the new class of applications that can be written, leveraging Hadoop. And so that's why I feel like it's definitely at the early stages. Jeff, what's your take on it? Well, I think, you know, the promise of Yarn is that you can now build new types of, or new ways of processing data so you can build real-time applications. You can use machine learning, you should graph analysis, streaming, all different types of ways to process data. But you still got to build the apps. So, you know, Yarn is, you know, it's yet another resource negotiator, but it's yet another enabler. It's not in and of itself the solution. You've still got to build the applications, find the right business case, you know, deliver applications that end users actually are going to use. So it's not a cure-all, it's a critical enabler. So my take is that, you know, it's going to take a little bit of time for that to happen. I mean, Yarn's only been here for six months or so, or maybe a little bit longer. You know, these things take time to develop. So it's fantastic that it's here, but I think it is going to take a while. But I fully think it's when those applications start to come online and that mainstream organizations are going to be much more likely to adopt some of these technologies. If you have to rely on internal developers to build your applications, you're not going to see the market kind of grow to the extent that we think it's going to. It's going to require some application companies to come out and, big vendors or specific startups to come out and develop these applications. I think you're going to see kind of a growth of new types of applications. Geospatial locations, so space and time dimensions to analytics is kind of really exciting to know people that are localized in certain areas. Are they surfing the same types of sites and do they frequent the same types of stores and are there certain affinities or understanding behavior while people are moving through time, right? Projecting and trajectory analysis of a person being and arriving at a certain spot at a certain timeframe or, you know, wearables are also increasing. The wearable conference is going on in San Francisco today. I have friends up there from Factual, Tyler Bell and the Factual team and our data sets is whole new internet of things. Absolutely, absolutely. I got to get your take on this. We were talking earlier with Merv about this one point and talking about some of these emerging startups. I won't say their name in San Francisco, but he said the quote, just because you put something in a bag doesn't make it a portfolio. Well said. Which is an old joke, right? It means you can't just because you put stuff into something doesn't mean it's actually cohesive. So I got to ask you the portfolio question. If you have the internet of things, you have big data. There are many, many, many, many use cases. So you guys have a portfolio. Some will argue, you know, some is better than nothing, some are emerging, but you guys have a comprehensive portfolio. Others are trying to replicate that. With all this diversity, how do you be successful as a portfolio business in the big data world? It's a tough question. So the big thing is really thinking about the client use cases and how the applications and the ecosystems actually evolving. So when you think about a portfolio, we really think about what I consider kind of the thread that weaves all these pieces together. And so for example, you know, streaming and operating in real time is a great need, but that in of itself isn't the outcome. So for us to think about the internet of things, for example, you know, just saying you've got a horizontal platform to handle that isn't enough, right? The internet of things means the data, the definition of the data set, how you process it, difference between managing it on the grid, managing it for deep well analysis, for oil rigs or for digging in coal and mining is very different than aerospace defense analysis or very different than connected vehicles and cars, which is very different than agriculture, which is different than cities and intelligent transportation. So when you think about what does it mean when we say the internet of things, there's a lot of machine generated data as well as human generated data and do you process it in a certain way that gets to that outcome. So for us, it's really important to understand the set of connectors, the way you're going to analyze and process that data, the full context, meaning all the associated information around that. So you have the context and in a perspective of dimension time entity as well as, you know, what are the data models? What are the algorithms? What are the unique customer experience? What's the level of the interaction you actually want to have? So I think about that full end to end dimension and quite honestly, there's a huge, huge team of experts. So it makes it easy. You guys have a lot of things going on but I watched like the cloud areas enterprise data how which I really love that positioning but it feels like they're pedaling as fast as they can and just so much energy when in reality you guys have the same concept with the hub but you have Watson is cognitive computing on top of it. So do you feel like you have an unfair advantage over the cloud areas of the world who are positioning themselves as a data hub when it's just a tub of data at this point, it's infrastructure? I'm not commenting. Okay. You know what it is. I mean, cognitive computing is essentially saying we're going to put reasoning on top of the data hub for extraction of real-time actual insights. That is the vision of what cloud area is saying. Well, you know, it's also kind of signaling kind of another shift that's happening on the technology dimension, right? So we went from what we call the tabulating era to the programmable era and a lot of advancements in programming languages has been really process driven. What we're saying is now programming languages are also evolving because of the nature of the data. Data is actually deciding what a process should be and what an application should look like which is fundamentally a different mindset. And then you get into cognitive which is really about some of the human elements like how do you interact? How do you perceive? How do you reason? How do you trust? How do you remember things? How do you learn things? It's a whole different advancement but I don't have quite the same view as the way you just described it because the way I think about it is it's an entire ecosystem of contributors that really enable the market shift and I consider it like, you know what, all boats rise when the tide rises, right? You know, but you want the tide to rise. You're being nice. We'll call it as it is. They're still early in your IBM 50 years of the mainframe under your belt. But the- I would say hands down, we understand the enterprise. There's no doubt about that. There's no doubt about that, but Jeff just did a segment with Jay from SSHB of Yahoo and some of the quotes coming in off the couch that were fantastic. Had Hortonworks stayed private, you wouldn't have seen all these other companies with tools coming out. So I bring this up with the cloud era question because their vision has always been this way from day one and even though they're small and growing compared to IBM, you guys have a dominant advantage over them, that competition is always good because no matter what if the cloud era wasn't pushing their agenda and Hortonworks didn't have their open source mojo, you wouldn't have all this new innovation. That's right, it's true, it's true. So do you feel that the innovation is hot right now? What would you look at on the landscape? What do you point at and say, that's hot innovation? What's the one thing that you point to right now that says that in your mind is the big innovation push? So first of all, I think right now data's sexy so pretty much anything in the data and analytics space is fairly hot. I think Forbes even announced last year that it was like one of the sexiest jobs to have if you had a data title, big data or analytics type title which I found interesting. Hey, storage was sexy once, now data's sexy, storage data, kissing cousins kind of thing? Who knows, who knows, it's all related I guess. But honestly it's like what clients get excited about when you can say it empowers kind of new work to be done, it empowers people to have access, it empowers a completely different interactive experience, it empowers a whole new level of insight and I think that's what drives innovation is it makes us pause and rethink about the possibilities. Rethink the possibilities on all dimensions, the business model, some of the disruptive business models are amazing, think about new roles and emerging roles that you didn't have some of these jobs 10 years ago or even five years ago or three years ago to things like operating in completely different ways that are going to be transformative for society, not just for business but really genuinely for people. Yeah, I mean the possibilities are just tremendous. I mean there is a kind of a journey that has to be taken, I mean there are going to be ways of doing things that are not going to happen anymore and that's going to impact the way people have to train to do their jobs and there's going to be much more data literate these days and it has the potential to leave some people behind but I fundamentally think that when we come out the other side the advances and the benefits to society and to business are going to be significant and this is not new, this happens with waves of innovation across history. Oh, absolutely, this is yet another one. But you know, it's the different types of partnerships like the work that we've done with the Mars family in terms of developing the cocoa beans for chocolate development because there was going to be a disease that was going to kind of eradicate some of the crop development. I mean looking at resource and water management, look at most of the urban populations where there isn't enough water usage or water, you know there's a restriction. Here, even in San Jose, right? There's dry seasons and there's different levels of pricing if you utilize and consume more than you're allocated, you think about rice grain development, you think about pollution, you think about transportation. I mean, huge, huge, huge macro global problems that we can begin to solve using data and analytics in a completely different way, so how cool is that? Well, great to see you on theCUBE, we're great to have you on there. I want to give you the final word, your thoughts around traveling the world and I know you said you're on a whirlwind tour. What's your big takeaway from going to Israel, going to Asia, circling the globe? What are you seeing? Big picture, macro trends right now in the enterprise, cloud, big data, consumerization of IT, Apple, iOS 8, yesterday they have some nice enterprise BYOD kind of features, I mean they're trying, right? Although Android's more buggy as some say, but with all that going on, what are the big trends that you see? Probably three things, so first is one that is huge is culture, so being much more analytic, data and analytic driven as an organization, so culture is one, doesn't matter what country, doesn't matter what language you speak, it doesn't matter if it's a secular or non-secular government institution, private institution. Second piece is investing in a platform that's much more current, leveraging capabilities like Hadoop, leveraging real-time context computing, some advanced analytics around machine learning, but that is, we really have to think about what is the next generation platform that's going to be ideal for your organization? And then the third is be serious about your investment into privacy and security. I mean, the amount of breaches, cyber attacks, but also the fraud aspects, it poses a whole different dimension that I would say in every country has come up very consistently. I understand it's like people will speak in a different language and all of a sudden I hear fraud, or da-da-da-da-da, cyber. I mean, it's like, you know, those three things I can understand. It's universal, right? Well, thanks for coming on theCUBE, great to see you. And again, so it's refreshing to see IBM really, you know, continuing its roots in open source, you guys have, you know, certainly in the generations going back, open source, been a great participant and player. And as we see in the Hadoop ecosystem, as Merve was saying and we were laughing about legacy, Hadoop will be the next buzzword. I forget the word you used. I forget the word you used this morning was, you know, what was the word? It was, that wasn't a legacy, that was what we're going to call it next year. Oh, okay. You don't remember? We don't remember having a blank. It's the end of the day. Now, check the crowd chat transcript. I'm sure it's under somewhere. Go check the crowd chat transcript. But Hadoop is growing up, business outcomes are being discussed. You're starting to hear the talk. It sounds like a market. It sounds like an immature market. You're hearing TAM, revenue, market share. We're not there yet, but we're starting to get there. Inheat you. Thank you for coming on theCUBE, IBM, big data. Here at Hadoop Summit, this is theCUBE. We'll be right back after this short break.