 5. From San Francisco. Extracting the signal from the noise. It's the Cube. Covering Oracle Open World 2015. Brought to you by Oracle. Now your hosts, John Furrier and George Gilbert. Okay, welcome back everyone. We are here live at Oracle Open World 2015. Also Java 1 happening today on Sunday on Howard Street in the middle of the road. This is the Cube. SiliconANGLE's flagship program where we go out to the events and extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE. George Gilbert, our Big Data Analyst at Wikibon Research. Part of SiliconANGLE. Our next guest is Neil Mendelson, Vice President of Big Data at Oracle. Cube alumni, great to see you again. Welcome back. Thank you. Good to see you. Good to be here. So obviously integrated cloud. Cloud is everything. End of the client server era. Welcome to the cloud era. We're here now. Last time we chatted, Big Data is pretty hot still. They don't have things. It's the hottest thing in the world right now. So they're interrelated. Continuing on the momentum of what customers are transforming to. The database is at the heart of all this. So we were talking to Tim earlier. What's your take on this? You've got the Big Data view inside Oracle. What's going on inside and what's presenting here at Oracle Open World? From a data point of view, data is expanding. The definition of data management has expanded beyond the database to include things like Hadoop and NoSQL. And we're embracing those things. And at the same time, we're talking a little earlier about how we're building a Big Data platform to allow people not only to ingest their data, but how they can run ETL on it, how they can visualize it, how they can get value out of it, run algorithms on it, and so forth. That's fun. So platform is a service for Big Data hot thing, you guys are talking about. But the number one question we get from customers that come on theCUBE is, look at bottom line, I need scale, I need speed, I need reliability, and it's got to be easy to operationalize. I need all those. What's your answer to that? What are you guys proposing? What are you guys seeing as successes? Can you share some insight? Yeah, I mean, the ease of operationalizing this kind of Hadoop stuff is really key to anything, right? I mean, it's key on the cloud, but it's also key on premises. And what we've been trying to do over the last few years is to really take all these disparate open source platforms, merge them together, put Oracle security on top, manageability on top, build products on top to really make it fundamentally much more straightforward for customers to essentially get value right away. So, Neil, if we talk about this Big Data platform, let's take the perspective that we're going to need packaged applications or at least semi-packaged applications for this to scale, because not everyone has the skills to build it. So, how would you define the extent of a Big Data platform? So, you know, we are starting to see third-party companies building applications on, you know, Hadoop, NoSQL, the database, the combination there is, right? We're beginning to see that in areas of customer 360, where apps are emerging there. We're starting to see it related to cybersecurity and other areas like that. So, the beginnings are starting, right, from that point of view. And along those lines, what assumptions do the developers make about what platform they're on? Do they say, oh, I just need to know that, you know, I have MapReduce or I have Spark and then I have, you know, Oracle for my operational data, you know, what are the components? So, you know, developers as well as customers are certainly free to operate at the lowest level. So, if they want to operate and write code on, you know, at the very low level, whether they're writing it in Spark or they're writing it in MapReduce, they certainly are free to do that. What we're trying to do is to up-level that, right, so that we can allow application developers, whether they're going through REST APIs or whether they're going through scripting languages, you know, it's always a question of, you know, if that consider programming or not, it doesn't matter, right? Or they can go through SQL and we're providing those data visualization layer, or data virtualization layer that allows them to abstract themselves from those direct technologies and allow them to be able to get one fast query over any of the data. So, just in terms of that data virtualization, you had talked about previously like polyglot approach, so that means you could take SQL and talk to the different capabilities underneath, whether it was, you know, machine learning or just executing a transaction and calling a propensity, say, to buy for a personalizing offer. That would be one way or all the interfaces supported in the Oracle Database, all the different ways someone might want to go. Yeah, so in the Oracle Database today, you know, we talk about either interfacing through SQL, right, interfacing through APIs, right, REST-based APIs, or accessing via scripting languages. Whether you access directly the Oracle Database or you have some data that exists in Hadoop. So let's take the example where you have web traffic data that's pooling into Hadoop. And then inside of the Oracle Database, you've got customer information. And the idea is you want to join those two together because there's nothing that says, what about the customer in terms of the clicks that are going on, right? And with Big Data SQL, the product that provides this virtualization, you're able to essentially perform that whether you go through SQL and API or REST-based interface. Yeah, talk about the competition for a second. Oh, that's up level, I'm a customer. Hey, you know, Oracle, Big Data Cloud. I'm really interested in an Oracle customer or I may not be an Oracle customer. I see Amazon's promoting a bunch of stuff. Azure's got a bunch of stuff. Now Hadoop, SQL on Hadoop, all this stuff's great. Bottom line, how do you compare versus Azure and Amazon specifically? So, you know, Amazon's got a great infrastructure from a hardware point of view, right? But beyond the infrastructure, they're not doing a whole lot, right? We talked about a moment ago, the ability to access across multiple independent data stores of different types, right? Not a capability that we have, right? You talk about, or not a capability that they have. Talk about other really strong capabilities. One of the ones that we recently started talking about was Big Data Discovery. This is a tremendous capability that allows you to visually see what's going on on your Hadoop system. So, when SQL's arrive on Hadoop, it synthesizes that metadata where no metadata exists and allows you to begin having literally a visual experience where traditionally that would have been writing code. So, if you're on Azure or for that matter on AWS, you're going to be writing code to find out what's on Hadoop. In our case, you're going to be using a search and a faceted navigation approach to be able to see the data rather than code. So, you have pre-existing software. Well, not just the developers, right? We're talking about business people, right? I mean, why should only developers have access to this stuff? So, I got to ask you. Last time we talked, you know, you have a history in the data warehousing business. It's well-documented in Oracle, billions and billions of dollars. You know, thanks to you, a little pat on the back for you, a plug. But that's one of the things that Amazon has actually highlighted with Redshift, that they're going after the low-hanging fruit with data warehouses. Now, the word Aurora comes out, that's the database piece. So, Amazon is going after Oracle, which is kind of like their early strikes, we bomb, we're still far behind Oracle. But what does that tell you about the customer environment? Data warehousing has got to be faster, smaller, cheaper. Kind of a Moore's law, if you will, or Allison's law from an Oracle standpoint. And now, Aurora, the database. Now, databases and big data obviously make sense coming together. What's your take on all of this and how you slice and dice that trend? First of all, when you look at the data warehousing market Oracle still has a commanding lead, continues to have that lead in the data warehousing space. But there's a lot of vendors that have essentially are zombies that are out there. IBM has done a good job of essentially assembling the zombies together. And they have Informix, they have Redbrick, they have Natesa, DB2 itself, largely given way to what they're doing with Watson. So I think that we're certainly picking up customers that are running on those zombie data-marts coming over to Oracle and perhaps Amazon will be able to do the same thing. But the likelihood that they're going to... That's not an innovation strategy. That's kind of keep the the zombies together, if you will. That's not real innovation. Well, I don't know, ask Amazon what they're... I don't particularly think migrating an old data-mart up to the cloud as being an innovation strategy, right? So what's your take on innovation? For the customer, hey, if thinking about the data warehouse of the future in an integrated cloud stack model, what is the nirvana for the future of data warehouses? Well, I think the data warehouse is now part of what we think of when we think of in terms of a big data platform. The data warehouse is an essential part along with the data lake and no sequel. So you're seeing those three environments essentially collapsing together to be able to provide a spectrum of capability. So this is key, though. Where... Who drives? Is it the transaction system that then, you know, calls to the analytics system to say, okay, tell me, you know, what I should offer? Tell me what I should do? Is there, like, with Oracle 12C, is that the layer that sort of orchestrates everything? Yeah, I mean, I think you're seeing kind of interesting mix-ups, you know, it's not only... It's not really what it used to be, such that you had your operational system standing by itself, you had maybe an operational data store, then you had a data warehouse, and then you had a bunch of people doing some, you know, we didn't call it data science, but they're writing algorithms somewhere in the background, right? And today, what we're seeing is the needs are driving customers toward wanting to have kind of a real-time basis, right? They want to be able to interact with the customer in real time, give them an offer, right, influenced by what they just did a few minutes ago. So, as an example, in the old world, you might have scored that customer giving them potentially an offer should they be interested in a credit card, right, and you've already determined what the risk profile of them based on their credit history and so forth. But maybe what you want to do today is you've got clicks streaming into Hadoop, and you want to influence that algorithm in real-time based upon perhaps the fact that they were on your website and then they left going to a competitor. You want to know that in real-time, right? So, the algorithm may suggest something different given a competitive situation than otherwise. So, we're seeing things kind of merge together and less separation between these systems and more of a need to act in real-time. Okay, so in Oracle's view, who's the main brain that's orchestrating that? Is it Oracle 12C? Or, you know, can it be Oracle sort of configured, you know, not as a transactional system but, you know, as a data warehouse where it could be either way? Yeah, I mean the orchestrator is always the developer, right, who puts the system together or the customer themselves and, you know, that might be coming from the Hadoop side, accessing Oracle, could be coming from a NoSQL perspective. It really shouldn't make any difference. So, what we're looking for then is, I mean, all the vendors are trying to bring that sort of the latency together, the speed, you know, accelerate what's coming in, making a decision, and then what's going out. So, as a result, what you're seeing is that, you know, in the old days what you had was these very, very, you know, I mean, Teradata was, you know, before that, Britain League, these systems that were really very batch-oriented that were intended for, you know, updates every quarter or every, you know, half year or something like that. And now what you see is the need for much more real-time support than you had before. Neil, open source communities have been a big part of the web. I'll say Tier 1, Resource Now, it's free, free software. I'll say Java 1's coming on. It says open source. How has Oracle embraced open source and share some specifics, but also share some insights as an industry vet around this disruption and this opportunity, because now you have internally Oracle, but your ecosystem is also embracing open source. Where do they come together? You know, we try to embrace it up and down the stack. So earlier I mentioned a new product called Big Data Discovery. Big Data Discovery is a native Hadoop tool, right? So it sits on top of Hadoop, right? In some cases it allows customers to very easily, as I've mentioned before, using machine learning algorithms to troll through the data, looking for the metadata and then being able to present it on the screen. And it's leveraging things like Spark and other technologies that are available from the open source community and doing that. So we're building production products based on open source stack while contributing to those stacks at the same time. What production products are they? All of them? Or pretty much all the ones in the Big Data stack, so whether it's Big Data Discovery or Golden Gate for Big Data, Golden Gate can now replicate in real time to do our Oracle Data Integrator product. At one point the Integrator would only be able to use the Oracle database to generate SQL or PL SQL to do the transformations. Now it generates Spark code, right? And you can run the transformations in Hadoop in addition or instead of running them on a more expensive machine. What are we expecting to hear this week at Oracle Open World for Big Data? What some of the highlights you could share now without kind of, you know, spilling the marbles in the lobby if you will share a little bit of color. I think that for me the most fun I think is really listening to customers tell their stories, right? So we've had a number of customers working on the cloud for some time on the Big Data space and we're going to be able to hear them talk about it in their sessions, right? One last question on this because the Oracle Cloud it works you know seamlessly replicates what you can do on premise but for customers who've made investments in, you know, Amazon and Azure when will sort of Oracle 12C running in a cluster working with Big Data appliance and Data Discovery, when will customers be able to, you know build their applications around that anchor on those other clouds? Well, I mean today Oracle is available on Amazon as well, right? Not really you can't really run it as a cluster it's true, there are different, yeah. I mean I think the, you know, the reality of it is that, you know, the world always looks simple, right? And you can just simply kind of plug all these pieces together, right? But things are not that simple, right? Things can be quite complex so, you know, within our own cloud what we're trying to do is to make it as seamless as possible and then to enable customers to be able to reach data from other clouds as an input or from an output as well and I think you'll see more of that. Talking, we were talking earlier about, you know, the Big Data sort of platform theme and many customers today who are grappling with that, the sort of build it yourself, they have a rich skill set in-house to do it. As it goes more mainstream should we expect to see more and more of these implementations to do for, you know, Oracle-led platforms for Big Data? Should we expect them to be more in the cloud? Well, we're certainly seeing, you know, and the DIY movement is really started at the whole beginning of this kind of open source to do kind of era and what we're seeing is customers that have large implementations, you know, over a thousand nodes, not in a single cluster now have whole hodgepodge of different clusters with different, you know, distributions with different tuned to different ways distributed in different ways and they're looking to standardize, right? We're also seeing small customers or smaller customers that are interested in getting into this, but they don't have those resources that you're talking about. What they want is to be able to take advantage of these technologies without having to figure out, number one, how to acquire all those people which are hard to come by, number one, and expensive number two, they just want to run the software, right? Or take advantage of it rather than put all the stuff together and I think that's what the cloud's all about. You know, here's some internal mantra that you guys are putting forth to your teams of big data. I see it's critical. Internet of things. I'm sure we're going to hear huge announcement around Internet of Things, big part of the big, big trend. Larry Ellison is no longer the CEO. He's the CTO, so I got to ask you, does he call you up and say, hey, we talk about this? I mean, how involved is Larry in a lot of these conversations? Can you share a little bit of color without getting bothered, without getting in trouble? I think Larry is as involved as he ever was, right? I mean, Larry was always... He gets down on the weeds, doesn't he? Larry was always keenly interested in the technology and he was and continues to be our visionary and I don't think that's changed, right? I mean, titles aside, right? I mean, you know, that hasn't changed. I mean, he's still very much involved in what we do on a strategic level and sometimes on a tactical level too. Nice smirk there. I said, I said two years ago when he shifted from CEO to CTO when he brought Mark Herden and software, CACB CEO. I said to Dave Vellante, I said, Dave, you watch. He's going to get down and the boat's going to get retooled. Now we're seeing some of the fruits of that labor. We saw last year's announcements. We saw the big data. We saw the cloud announcement that Larry gave the keynote where we interviewed you last. A lot's happening super fast. This is the boat race. This is Larry's boat race. How fast is it moving in your mind right now? I mean, obviously you're seeing Dell Oracle directly. It's exciting. I'm sure Larry's getting down there. How fast is it moving? There's nothing new with companies going after Oracle directly. We've had that nearly for the beginning. People love to do that and often they just bounce off or they kill themselves in the process somehow. I think software talked about that just the other day when being interviewed. There's nothing new there. The more the market gets, the more innovations coming from different places, the more it really invigorates the culture of Oracle to do more, to do faster and to do better. In a hyper-competitive environment is the environment where we absolutely thrive. People bring their A-game. People always brought their A-game. First day at Oracle, revelation of myself was like holy crap, everybody here is the smartest guy in the world. That part hasn't changed. We love talking to you guys. I got to ask the final question. We're going to have John Falleron later today at 3 o'clock. Unix is back. It never left. Same with Java. Steve Jobs did this in the industry when he took over the Mac. Next in became the bulletproof operating system. I got to ask you in the big data world where cloud and performance are powering analytics. All kinds of analytics from a wide range of variables. How important now are the engineering systems? Specifically, what are you most excited about that's coming out that's going to give major lift to the big data space? Well, the engineering systems are important to us. Both in terms of being able to achieve very high performance, but also high availability. The fact that it's a standardized platform means that whatever issues that we come across we come across them in thousands of customers. We can get that value from an availability point of view across that and the speed that comes with it. We're looking at all kinds of different things coming down the pike, some of which from Intel, some of which innovations coming on our own side, so it's an exciting time. Neil, always pleasure to have you on theCUBE. You're a content machine. We love having you. You're so smart and great, good color. Final question, final, final question. Put the bumper sticker on this show early. I know there's a lot coming, but Sunday people want to taste what's the say on the bumper sticker for Oracle Open World this year? Wow. That's it. Big fun? Yeah, it's a big fun. Neil Mendelsohn, Vice President of Big Data Group here at Oracle Open World. This is theCUBE, live on Howard Street, and we've got more guests coming. John Fowler, the top executives. We're going to do 40 interviews throughout Wednesday, a few live here on Howard Street, Oracle Open World 2015. We'll be right back with more after this short break.