 DataWorks Summit Europe 2017, brought to you by Hortonworks. Hello everyone, welcome to theCUBE special presentation here in Munich, Germany for DataWorks Summit 2017. This is the Hadoop Summit powered by Hortonworks. This is their event and again shows the transition from the Hadoop world to the big data world. I'm John Furrier, my co-host Dave Vellante. Good to see you, good to see you Dave. We're back in the seat together, usually on different events but now here together in Munich, great beer, great scene here. Small European event for Hortonworks and the ecosystem but it's called DataWorks 2017. Strata Hadoop is calling themselves Strata and Data. They're starting to see the word Hadoop being sunsetted from these events which is a big theme of this year. The transition from Hadoop being the branded category to data. Well, you're certainly seeing that in a number of ways that the titles of these events, but first of all, I love being in Europe. These venues are great, right? They're so Euro, very clean and magnificent but back to your point, you're seeing the Hadoop Summit now called the DataWorks Summit. You've seen Strata plus Hadoop is now Strata plus, I don't even know what it is, right? It's not Hadoop driven anymore. You see it also in Cloudera's IPO. They didn't talk about Hadoop and Hadoop Distro. They're Hadoop Distro vendor, but they talked about being a data management company and John, I think we are entering the era or well deep into the era of what I have been calling for the last couple of years, Profitless Prosperity. Really, when you see the Cloudera IPO, as you know, they raised money from Intel over $600 million at a $4.1 billion valuation. The Wall Street Journal says they'll have a tough time getting a billion dollar valuation. For every dollar each of these companies spends, Hortonworks and Cloudera, they lose between $1.70 and $2.50, so we've always said at Silicon Angle, Wikibon and theCUBE that people are going to make money in big data, other practitioners have big data and it's hard to find those guys, it's hard to see them, but that's really what's happening is the industries are transforming and those are the guys that are putting money into their bottom line, not so much the technology vendors. Great to unpack that, but first of all, I want to just say congratulations to Wikibon for getting it right, again, as usual, Wikibon ahead of the curve and getting out there and getting it right because I think you're nailed it and I think Wikibon saw this first of all the research firms kind of pat ourselves in the back here, but the truth is the practitioners are making the money and I think you're going to see more of that. In fact, last night, as having a nice beer here in Germany, I just like to listen to the conversations in the bar area and a lot of conversations around, real conversations around doing deals and deployments. You're hearing about HBase, you're hearing about clusters, you're hearing about service revenue and I think this is the focus. Cloudera, I think in a classic Silicon Valley way, their hubris was tempered by their lack of scale. I mean, they didn't really blow it out, I mean, now they do 200 million in revenue, nothing to shake a stick at, they did a great job, but they're buying revenue and Hortonworks is as well. But the ecosystem is the factor and this is the wild card out of making a prediction. Profitless prosperity that you point out is right, but I think that it has longevity with these companies like Hortonworks and Cloudera and others like MapR because the ecosystems robust. If you factor in the ecosystem revenue, that is enough rising tide in my opinion. The question is how do they become sustainable as a standalone venture that Red Hat for Hadoop never worked as Pat Gelsinger predicted. So I think you're going to see a quick shift and pivot quickly by Hortonworks. Certainly, Cloudera is going to be under the microscope once they go public. I'm expecting that valuation to plummet like a rock. They're going to go public, Silicon Valley people are going to get their exits. But Mark- Excel will be happy. Everyone's going to, yeah, they'll be happy. They're already sold in 2015. They did a big sale. I mean, a lot of them cashed out in two years ago when that liquidation event happened with Intel, but that's fine. But now it's back to business building and Hortonworks has been doing it for years. So, well, we know- And their valuation is less than a billion. So I'm expecting Cloudera to plummet like a rock. I would not buy the IPO at all because I think it's going to go well under a billion dollars. Well, that is just the right call. And as we know, last year, then the last year, Fidelity and other mutual funds devalued their holdings in Cloudera. And so, you've got this situation where, as you say, there are a couple hundred, maybe on their way to 300 million in revenue, Hortonworks on their way to 200 million in revenue, add up the ecosystem, maybe you get to a billion, throw in all of what IBM and Oracle call big data and it's kind of a more interesting business, but you've called it same wine, new bottle. Is it a new bottle? Now, what I mean by that is the shift from Hadoop and then, again, you read Cloudera's S1, it's all about AI, machine learning, the cloud, interesting, we'll talk about the cloud a little later, but is it same wine, new bottle, or is this really a shift toward a new era of innovation? It's not a new shift, it's the same innovation that the Hortonworks was founded on. Big data is a categorical and Hadoop was the horse they rode on, but I think what's changing is the fact that customers are now putting real projects on the table and the scrutiny around those projects have to produce value and the value comes down to total cost of ownership and business value and that's becoming a data specific thing. You look at all the successes in the big data world, Spark and others, you're seeing a focus on cloud integration and real-time workloads. These are real projects, this isn't fantasy, this isn't hype, this isn't early adopter, these are real companies saying we are moving to a new paradigm of digital transforming our companies and we need cost efficiencies but revenue producing applications and workloads that are going to be running in the cloud with data at the heart of it. So this is a customer forcing function where the customers are generally excited about machine learning. Moving to real-time classification of workloads, this is the deal and no hubris, no technology posturing, no open standards, jockeying, can write the situation. Customers have demands and they want them filled and we're going to have a lot of guests on here and I'm going to ask them those direct questions. What are you looking for? Well I totally agree with what you're saying and when we first met it was right around the midpoint of the web 2.0 era and I remember Tim Berners-Lee commenting of all this excitement everybody's doing he said this is what the web was invented to do and this is what big data was invented to do. It was to produce deep analytics, deep learning, machine learning, cognitive as IBM likes to brand that and so it really is the next era even though people don't like to use the term big data anymore. We were talking to some of the folks in our community earlier, John, you and I, about some of the challenges. Why is it profitless? Why is there so much growth but it's no profit? And we have to point out here that people like Hortonworks and Cloudera they've made some big bets. Take HDFS for example and now you have the cloud guys particularly Amazon coming in with S3. Look at Yarn, big open source project but you've got Docker and Kubernetes seem to be mopping that up. Tez was supposed to replace MapReduce and now you've got... I mean I wouldn't say mopping up, I mean you've got Spark. At the end of the day the ecosystem is going to revolve around what customers want and portability of workloads, Kubernetes and microservices, these are areas that just absolutely make a lot of sense and I think people will move to where the frictionless action is and that's going to happen with Kubernetes and containers and microservices but that just speaks to the DevOps culture and I think Hadoop ecosystem again was grounded in the DevOps culture. So yeah there's some projects that are going to have maybe go out of flavor but there's other stuff coming up through the ranks in open source that I think is compelling. But where I disagree with what you're saying is the point I'm trying to make is you have to, if you're CloudEra and Hortonworks you have to support those multiple projects and it's expensive as hell. Whereas the Cloud guys put all their wood behind one arrow to use an old Scott McNeely phrase and Amazon I would argue is mopping up in big data. I think the Cloud guys, it's ironic to me that CloudEra in the Cloud era picked that name but really never had... They missed the Cloud. Really they've had a strong Cloud play and I would say the same thing with Hortonworks and MapR. They have to play in the Cloud and they talk about Cloud but they've got to support Hybrid, they've got to support on-prem, they've got to pick the Clouds that they're going to support AWS, Azure, maybe IBM's Cloud. Look at CloudEra completely missed the Cloud era pun intended. However they didn't miss open source. What they're great at and what I'm an admirer of CloudEra and Hortonworks on is that their open source ethos is what drove them and so they kind of got isolated with some of their product decisions but that's not a bad thing. I mean ultimately I'm really bullish on CloudEra and Hortonworks because of the ecosystem points I mentioned earlier. I'm not high on the I wouldn't buy the IPX I think I'd buy that at a discount but CloudEra is not going to go away Dave. They're going to go public. I think the valuations going to drop like a rock and then settle around a billion but they have good management. They have a good, the founders still there, Mike Olson, Amar Awadala. So you're going to see CloudEra transform as a company. They have to do business out in the open and they're not afraid to obviously they're open source. So we're going to start to see that transition from a private venture backed scale up by revenue, the playbook of Silicon Valley, venture capitalists, Excel partners and Greylock. Now they go public and get liquid and then now next phase of their journey is going to be build a public company and I think that they will do a good job doing it and I'm not down on them at all for that and I think it's just going to be a transition. Well they're going to raise what? A couple of hundred million dollars but this industry. Two hundred million. Yeah this industry's cash flow negative. So I agree with you, open source is great. Let's rah-rah for open source and it drives innovation but how does this industry pay for itself? That's what I want to know. So how do you respond to that? Well I think they have sustainable issues around services and I think partnering with the big companies like Intel that have professional services might help them on that front but Mike Olson said in his founders letter in his S1 kind of AI washing, he said AI and cognitive. Totally AI washing. But that's okay because CloudEra could easily pivot with their brain power and same with Hortonworks, to AI. Machine learning is very open source driven. Open source culture is growing, it's not going away. So I think CloudEra is in a very good position. I think the Cloud guys are going to kill them in that game. Cloud guys on IBM are going to cream these profitless startups in that AI and machine learning. Well we'll see. You disagree? I disagree, I think, well, I mean it depends. I mean, you know, I'm not going to forecast what the management might do but I mean if I'm looking at what CloudEra's done. No, I would do it exactly what Mike Olson's doing is I'd basically pivot immediately to machine learning. Look at Google, TensorFlow has got so much traction with their cloud because it's got machine learning built into it. Open source is where the action is and that's where you could do a lot of good work and use it as an advantage in that they know that game. So how do they make money at that? I would not count out the open source game. So we know how IBM makes money at that, you know, in theory, anyway, Watson. We know how Amazon's going to make money at that with their proprietary approach. Microsoft would do the same thing. How do CloudEra and Hortonworks make money? I think there's a product transition around getting to the open source with cloud technologies. Amazon is not out to kill open source so I think there's an opportunity to wedge in a position there and so they just got to move quickly. If they don't make these decisions, then that's a failed execution on the management team at CloudEra and Hortonworks and I think they're on it. So we'll keep an eye on that. No, Amazon's not trying to kill open source, I would agree, but they are bow guarding open source in a big way and profitingly. No, they get customers using it. Profiting amazingly from it. Well, they just do what, as Andy Jassy would say, they're customer driven. So if a customer doesn't want to do five things to do one thing, this is back to my point. The customers want real time workloads, they want it with open source and they don't want all these steps and the cost of ownership. That's why this is not a new shift. It's the same wine, new bottle because now you're just seeing real projects that are demanding, successful, and efficient code and support and whoever delivers it builds a better mousetrap. In this case, the better mousetrap will win. And I'm arguing that the better mousetrap and the better marginal economics, I know I'm like a broken record on this, but if I take Kinesis and DynamoDB and Redshift and wrap it into my big data play, offer it as a service with a set of APIs on the cloud like AWS is going to do or is doing and Azure is doing, that's a better business model than five different pieces that I have to cobble together. It's just not economically viable for customers to do that. Well, we've got some big news coming up here. We're going to have two days of wall-to-wall coverage of DataWorks 2017. Hortonworks announcing 2.6 of their Hadoop data platform, Hortonworks data platform. We're going to have Scott now, the CTO coming up shortly. Stay with us for exclusive coverage of DataWorks in Munich, Germany, 2017. We'll be back with more after this short break.