 Live from New York, it's theCUBE covering big data NYC 2015. Brought to you by Hortonworks, IBM, EMC, and Pivotal. Now your host, John Furrier at George Gilbert. Okay, welcome back everyone. We are live in New York City with SiliconANGLES theCUBE, our flagship program. We go out for the events and extract the signal and noise. I'm John Furrier, founder of SiliconANGEL, my co-host, George Gilbert, Wikibon's big data analyst. Our next guest is CUBE alumni, Lawrence Schwartz, Chief Marketing Officer, Attunity. Welcome back, it's like our sixth year you've been on theCUBE, congratulations. You've been in every, it's like, it's like you're a guest analyst now. We got to keep you into theCUBE. Always great perspective, good to see you. Always a pleasure to be on here and hear what's going on. And we have a little surprise, a little virtual reality at the end. So stay, keep watching. You'll see what we're going to show at the end. Great, great Attunity traction over the years. You guys have been successful. And you've been swimming in the sea of noise out there. Obviously the Hadoob ecosystem has been very noisy. Certainly the growth is there. A lot of new announcements, a lot of new innovations. But now the theme here this week is the rubber hitting the road, put some meat on the bone, Michael Olson's up on stage, the keynote saying, it's about the outcomes and solutions, which is like code word for, you better start booking revenue guys, or you do not have value in your products. Sure, absolutely. So that's what we're hearing from across the board, unanimous that it's time to make some money. That's a proxy for your value. You guys have been doing some successful work with customers. So everyone wants to know, not on the geek side, but on the practitioner, I am going to deploy Hadoop in conjunction with other ecosystem stuff. Exactly, yeah. What is the story? What's going on? What is on the top conversations for the enterprise customers? Sure, so what we focus on at Attunity is helping customers in a number of areas that really help folks in the Hadoob realm. One is you got to think about your legacy systems, your systems of record, all these existing Oracle databases and data warehouses, all the existing stuff you might have on Teradata, all these other systems that are not going anywhere anytime soon. They're going to be a part of the new ecosystem. So you have to think about how you combine that, how do you get, I think as George has written, the systems of intelligence, and get more value out of it. So customers are thinking about that as they think about how to deploy Hadoop. So we have solutions that help you kind of mix the two, kind of come up with the optimal solutions, both for balancing where you put the workloads, figuring out what makes sense to go where. In the process that enables customers, we know we have lots of data, migrate over to parts of Hadoop where they want to and where they need to, and so we give them solutions for that. And it's not just small test cases. We've seen that over the past years and as you were mentioning, we've really seen kind of the need for the enterprise solutions. And when I talk about the enterprise solutions, just to give you some examples, we have customers who, large bank where they've got a large existing data warehouse. And they know that they want to kind of cap the growth of that. They don't want that to grow anymore, but they still have more information coming into it. So they're trying to figure out, how can I keep that at a certain level and then what do I start building out on Hadoop to take advantage of some of the cost points. So for that, we can help customers go in there, look inside their actual data warehouse and figure out what's really being used, how often and when, kind of look for the hot and cold date, if you will. And we give really good visibility with our so-called visibility product on what's happening to the different business lines as well. So you can see a finance is claiming they're going in there and using it. Are they really fully using it? Marketing, other departments, production. So you have that really good insight into what's happening in your current environment and then you can make really intelligent decisions about what do I want to offload and whatnot. So that's one area. And then we've continued to expand that solution set. Now we've gotten to the point, we have some customers who have put so much on Hadoop. We have a customer now who has six petabytes on Hadoop, major online travel company. At that point, you start looking at this and this can't just be a single data lake for a customer. You have to start thinking about this like you would with any type of data warehouse or enterprise environment. How do I start cheering that? How do I put the more valuable workloads on some processors that have more compute power, some nodes that have more compute power, some faster storage, and set up several tiers, maybe archival tiers. So we've expanded our solutions and that's one of the things we're talking about at the show is we've taken that visibility product which gave people great insight into their existing data warehouses and add that capability to large Hadoop data lakes and going in there and figuring out what's going on. Okay, so this is kind of interesting in the sense that we know running costs, we know capital costs for the Hadoop platform per terabyte and for the data warehouses per terabyte, roughly on the order of maybe 2,000 a terabyte on the Hadoop side up to 25, 35,000 a terabyte on the data warehouse. Have you been able to guess or estimate the running costs of the two platforms with all the admin mixed in? Yeah, we certainly see those 10 to one disparities that you've mentioned. I think from what we hear from customers, oftentimes it could be smaller than that. When you add in all the training, all the operational support, it might be a couple of X or four or five X or something like that. So in practice it might not be so dramatic but once they have the skill set and once they've kind of gotten over the Hadoop pump if you will, then that becomes easier to deal with. Okay, so talk about the buyer side of it. I want to get back into the consumption. We always talk about this in the cloud business. Pretty easy now to kind of get a handle on the directional path for cloud. Public cloud, Amazon, Azure's trying to put some stuff up there, that kind of has a hybrid cloud. Private cloud and on premise, and then hybrid is the engineering solution of both. Is there some clarity around big data consumption that you're seeing with customers? And can you be specific on how they're buying? Sure, sure. Well we see, when you look at the cloud and we have a lot of customers who do that as well, we have strong partnerships with AWS and Azure and whatnot. So we see that the buying behavior and what they're using with the cloud for is along the lines of they want to kind of spin up some easy to run analytics, right? So like on AWS it might be Redshift and it might start at the department level, right? So there's a much lower barrier to entry from some smaller department that just wants to start and get going and that's where they'll kind of start looking at using that. Oftentimes that's good enough, they'll stay there but then they look at the economics of scaling that out and the hourly cost and if it gets really big then it's a decision of hey, I might want to bring this on-prem. But we see that need for a self-serve, right? Kind of starting up in different departments and then getting going. We saw that first with our cloud customers and that's the other thing that we're now enabling too with our replicate solution which lets people move over data and keep it in sync. The other thing we're announcing at the show is we have an express version of it so that lets customers in a department start easily, download it, get going and then have alternatives particularly in the Hadoop world to scoop, to get going and so we're trying to enable those types of uses. So what's your take on the overall ecosystem? What's your personal, it's an industry participant put your analyst hat on, not your attunity hat. What is your reaction to what's happening in this year's show? Sure, well I see the show is just starting up. We see a lot of noise, kind of some of the announcements that are coming out now. So Hadoop, I think you were talking about it earlier, you see the enterprise, how do you make that an enterprise version? I think people are also starting to talk more about how to stream in different sources. The internet things people have been talking about for a while but I think you're seeing more and more solutions on that and people are going beyond the typical ways they pull in data into a data warehouse. They're trying to figure out how to make things more real-time in Hadoop or leveraging Spark so we kind of see that happening so I think there's going to be some excitement there over the years but at the end of the day these systems of record, if you will, are not going anywhere so people always have to figure out how to coexist with them and that's where I think some of the opportunities, at least opportunities for companies like us is you've really got to figure out how to merge what you have with what's going forward. The way you position systems of record is not really going anywhere is interesting because we see that as the plumbing for the systems of intelligence. Absolutely. The question I have is do you have to do a different performance profiling for the systems of record to understand or at least the data warehouse or the analytic part of it to see that they perform to the level that's required for the systems of intelligence? Yeah, no, you absolutely. The first part of any of this when you migrate over turn into a modern data architect and you're having kind of a mixed system like this is you've really got to assess what's going on and what's working. So we work with some of the big system integrators and that's the first part of their services is an assessment that they do of their existing environments to kind of see what's going on, where the opportunities are, what's working, what's not. So that's absolutely kind of a precursor to it. And we do that, we do that some interesting ways in the technology. You're looking at all the queries, all the asks that go into a data warehouse and you're doing a parsing of that and semantic parsing to really understand how's it being used, under what circumstances and even before you change it you might go in there and see that boy, some of these queries aren't optimized, they're taking up a lot of resources. Even before you add anything, there's a whole lot of homework you might do just to kind of optimize your existing performance. And then you can at that point figure out do I need to optimize it further? Do I need to move it somewhere else? Some people might move it from an existing data warehouse to another tier data warehouse or they might then move it to Hadoop. And then as I mentioned now, we're getting to that extreme case of petabytes on Hadoop and you got to figure out how to manage that as well. So what's the real reason why customers are buying from you guys? Why are you guys winning? Can you narrow it down to some specifics, simple example or in common? Absolutely, yeah. So it's a few things that really kind of help us help our customers. One is if you look at what people do today like in the Hadoop world to kind of get started going you've got to get the data in from somewhere, right? So they might use a tool like Scoop, right? To get started and move it in. And that's a very developer driven, set up for a sandbox environment, difficult to use, doesn't keep things in sync well, right? So there's kind of some initial tools that people can use to kind of get started, but they're not really enough to be, you know the enterprise level that we're talking about, right? For an enterprise wide deployment. So they're looking, they crave these solutions that are much easier to use to really handle their data for batch loading, for change data capture. And that's where Replicate comes in and really differentiates and sets itself apart. Very easy use, very good to get going. And that's been the case when we do it for databases, data warehouses on the cloud and you know, in the last few years with Hadoop as well. So that's one area is making that real pull of fruit. And we see customers who otherwise, without a tuning they would take, you know, DBA's worth of time for a couple of weeks or months. They come in, they can use Replicate and do this all within a few hours. So that's one area. And then the other is, as we were talking about earlier just really getting that visibility and what's going on to environments. A lot of the tools that give you visibility into a data warehouse or Hadoop environment they're really geared towards the DBA, right? They're kind of technical tools. They don't, you know, and they're really for debugging kind of performance. We really give you a business wide view across the entire data warehouse. So business user can come in and see what's happening, you know, for their environment, where the issues are, where the growth is, where the opportunities to move are. And we make it very easy to kind of dive in and out graphically and see what's happening. You know, we just completed a very comprehensive survey of 300 practitioners. Just got the results back last week. We're still sort of, we've tabulated them but we've still got to organize all the answers. One of the things we found out was that for a majority of the respondents, the data warehouse certainly wasn't going away. It was sort of being maintained roughly at its current size. And Hadoop was, you know, capturing some but not all the growth. Is that what you're seeing? We're seeing some customers who definitely want to do that, right? They have a large instance of a data warehouse and they want to kind of cap, you know, the growth of that and look for alternatives. And we see that, you know, across our environments. We see that, you know, for a couple of different data warehouses. We even see people talking about that we have solutions for even SAP and people think of that in the same way about how do I cap my SAP growth? So yeah, that's absolutely, people are looking for ways to cap what they have and then how do they build out what the new solutions that give them different points. And when they're capping their SAP growth, are you talking about the, because in their case, Hanna started out as underneath their data warehouse, but now they're trying to essentially combine the operational system and the analytics in one database. Are you monitoring performance on that? Because that's an interesting use case where the world is moving towards integration of the operational app and the analytics. And I'm curious if that's sort of a canary in the coal mine, what does that look like and are you seeing it elsewhere? So what we've seen on the SAP side, we have a different set of solutions, one called Go Client, and what helps you manage the SAP landscape. And what we started doing there was we have a relationship with HP to help people archive off of that. So that becomes an absolute use case and that's where you talk about large systems that they realize take a bank that has to do a seven year plus archive of all their data, they don't necessarily need to keep that all on SAP, they might need to keep the archive there, they might need to access it at some point, but if we can help pull that information off, in their case, put it to an HP archive, that works, that meets the regulatory requirements, it kind of sets up a different environment. So conceptually, it's similar to what we see on some of the data warehouse in Hadoop sides. Okay, got to ask you, we are in our virtual reality space here. What are these things you're giving away? I want to show the folks out there. This is an Intunity branded Oculus Rift cardboard version, iPhone in the inside, which is really cool. So I'm looking at this, and I'm pulling around, there's a little thing in the middle there, and it says I can pull the trigger on the start button. They just stare at the start button and you're on your journey. So what is this? Are you guys giving these away to booth? What's going on with these guys? Yeah, yeah, come by the booth that we've got these great little Google cardboard apps here. You can use your own phone and Android, there we go, Android. And you've got iPhone as well, either one, just download the app from the store. And it's a lot of fun, you can go in there. What we're trying to show is here, look, people have large data lakes are often becoming a data swamp, right? So how do you get visibility into that? How do you get clarity out of it? So this takes you on the viewpoint of the data, going through a journey, migrating over to where it needs to go, all the issues you have come across. And it's kind of fun and entertaining. So come by, we've got these to give out. We have it online as well, you can download the app and have some fun with the community. It's all about having fun with data. And just so give the plug forward, what's going on with the booth down at the show? Absolutely, yeah, yeah. So we've got a booth over at the show, we'll be there for the whole experience. Later today, we have one of my colleagues, Rodan, giving a talk as well, talking about the different things you need to think about when managing and moving data. So he'll be giving a talk. But yeah, we'd love to hear you come by and tell us what's going on. Actunity inside theCUBE and remember tonight at seven o'clock Eastern, you're going to see, I mean, five o'clock Eastern from five to seven, live presentation, George, we give you his research and party after. More than welcome to come by, I appreciate it. We are live here in New York City for Strata plus a dupe and our event, Big Data NYC here, one block, 100 yards from the Javits Center. We're on the ground, getting all the data and sharing with you as real time in the moment live. We'll be right back after this short break.