 Live from Boston, Massachusetts, it's The Cube at the HP Vertica Big Data Conference 2014. Brought to you by HP. With your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. Here we are live in Boston, Massachusetts. This is The Cube, SiliconANGLE, and LukiBond's flagship program. We go out to the events. I'm John Furrier, founder of SiliconANGLE. My co-host through the segment, Jeff Kelly, LukiBond senior analyst. Our next guest is Josh Rogers, president of SyncSword. Welcome back to The Cube. Great to see you. Great to see you. We get to see you here at all the big data events. You guys are becoming a household name in the big data industry. And the fact that you have so much experience and track record, it's great to see you guys have such great success. Congratulations. But we're here at the HP Vertica Big Data Conference. So tell us what's going on with you guys and we've got this Hadoop world coming upon us very fast. Vertica fits well into that world. We had just had some great conversations earlier. How do you guys fit into that ecosystem with Vertica, the splunks of the world, everyone else? Yeah, sure. So we've actually had a long-standing relationship with Vertica really since they launched the original year of the product and have optimized our offering to be able to be the fastest load mechanism into Vertica. So we have most recently updated that towards the end of Q2 and built an intelligent connector that does a few different things that will help customers move data more seamlessly into Vertica environments. The first thing is we've created a highly parallelized connector and what it does is it automatically determines how many parallel connections to establish to the database directly to the nodes and then it will actually balance the workload as the load happens, depending on which nodes can accept data the fastest. And what we're seeing is customers are really amazed at the power of Vertica, but it's obviously limited by how much data they have in their instance and so the ability to kind of load large volumes of data in a not only a high capacity manner, but an intelligent manner that makes efficient use of resources is something that they value highly. In addition, one other point, you can do that directly from Hadoop and using reducers, leveraging DMX age. So I got to ask you, what solution don't you have? You guys have mainframe, you have the cloud, you have industry solutions. It's interesting, right? Some people, Dave and I always speculate the cloud is essentially big iron that's decentralized called the mainframe in the cloud. We're kind of coming back to that world, right? So you've been there, done that, synced to it as a company and have all that experience. You guys are in a good position. Are you worried that you're not focused enough? Let me simplify it a little bit. If you look at the core IP of what we have, it's an ability to take the processing of data, set based operations on data, and do that in a highly performant manner, but it also leveraging efficiency and intelligence to make that performance easy to get and to make efficient use of resources. And that's true on the mainframe in 1968, and it's true on our open systems products, starting in the 90s. And now what we're doing is going around the ecosystem of big data and making it easy for customers to tap into those processing capabilities, regardless of what kind of other technologies they're using, Vertica, Hadoop, most recently renounced the relationship with Splunk. That's a really interesting one. Now what we're doing in dot-conference is the cube will be there at Splunk. So Splunk obviously went public, great success. I got to ask you though, because this is something that Jeff and I were talking about prior to this event, and with Dave Vellante, and certainly our team in Silicon Valley is, the cloud has opened up the developer market and changed the game on the developers. So you're seeing DevOps as a major driver for innovation, and you've got categorically two types of developers. Born on the cloud, the young guys, we would never load Linux, they'd just run the stack from the cloud, and the old guys like us, I guess me, and then anyone over pretty much 30 has load Linux and spin around, open source certainly standard across. So old school, and then new schools born in the cloud, that's a huge development change. What's your take on the developer market? How do you guys talk to the young developers, or do you, and how do enterprises bring some of that mojo of DevOps into the enterprise? Yeah, sure. So we certainly talk to the young developers, we have customers that span both large enterprise traditional industries as well as next generation companies, whether it's e-commerce or gaming, and so we have developers using our tools from both those pockets of the community. Again, our strategy is to make our software available for how they'd like to consume it. So we have, in November of last year, we launched the Hadoop offering in the cloud. In May of this year, we offered the ETL version directly on EC2, and so what we found is huge uptake by the development community to build integrations in the cloud. Some of those get run in the cloud, some of those get run on premise. We had a customer in Q2 that was starting on premise and using our technology to develop their data integration routines, but their ultimate goal is to run everything in the cloud. So let's take into that a little bit more. So we certainly hear about cloud and big data kind of mentioned in the same breadth a lot, but kind of cloud, one that other was concerned about security. I think security concerns are starting to fade a little bit. When you talk about Amazon, they're just as secure as any on premise data center. But from your vantage point, you have a good vantage point because you're looking at clients that are both running on premise and that are thinking about moving to the cloud, as you just mentioned. What are the dynamics that you're seeing in that environment? Are you starting to see a migration towards the cloud for beyond just kind of POCs? Is it becoming a destination for production workloads? Yeah, for sure. I think one kind of construct that's probably safe to assume is that everybody's got a mixture of workloads, some on premise, some in the cloud. Some are just starting in the cloud and it's just dev, but their goal is to move that more aggressively for actual production workloads in the cloud. And it's getting over some of these security hurdles and some network bottlenecks, et cetera. But that's the direction that people are moving. And sometimes that's a private cloud, sometimes that's an Amazon, and I think we're going to be seeing people take more and more advantage of public cloud infrastructure, particularly as security gets more sophisticated. The economics are, it's hard to argue with the economic. So is that largely the economic question, or the economic factors why people are looking to make this move or is it a combination of that and the technical complexity of running some of these big data systems on premise? I think it boils down to economics, to get the level of flexibility to be able to scale up and scale down workloads, the amount of infrastructure you have to build in an on-premise fashion just becomes economically unfeasible. So at the end of the day, I think it's economics, but there's certainly a level of agility that people get that causes them to pursue that strategy. But everybody at this point has some piece of their portfolio, whether it's dev, whether it's some subset of applications that are running in the cloud. Well, it's interesting because if you talk to some CAOs, they say, no, no, we're not using the cloud for anything. And then you talk to some of the developers, you're talking about the application or that application. And sometimes there's not that visibility from a high level, from a CIO level, which could create complications of its own. Well, that's true in Hadoop as well, right? Oh, yeah. There's lots of CIOs that don't know how much they're doing with Hadoop because it's been started by shadow IT groups on the business, et cetera. Yeah, with Hadoop, we hear that with NoSQL all the time. We're not using Mongo for anything. Well, actually, yeah, you've got some developers using it for a number of things. Interesting. So let's dig into the Splunk relationship because obviously they're one of the hottest, big data companies out there, and at least in terms of the public markets, they get a lot of attention from the mainstream press. So what are you guys doing with them? Yeah, so as you know, Splunk allows you to kind of glean operational intelligence out of machine data. And they've got this vertical stack that allows you to take log data and visualize it and understand aspects of either your operational environment, risk, fraud, et cetera. And there's a bunch of apps that people have started to build on top of Splunk. And one of the most pervasive platforms on the planet, particularly in the enterprise, is the mainframe. And what people may not know about the mainframe is it probably has the richest log data of any system on the planet. So the challenge is that that log data needs to be intercepted and moved into a Splunk environment in the right way that doesn't put a big load from a CPU perspective on the box. The data formats are very different. There's actually a tremendous number of types of logs. So there's this notion of SMF records, which there's 100, and I think there's 200 different types of SMF record with each one having some components. So being able to kind of interpret these logs is fairly complicated. What we've seen consistently from Splunk's customers is that they are looking to build scenarios that incorporate both open systems log data from, you know, Linux environments, et cetera, with their mainframe log data. Today they don't have a good way, either cost-effective or technically, to get that log data on a time basis into their Splunk environments. Splunk doesn't have residents in its development group the expertise to be able to decipher that mainframe log data. We obviously have that. So what we're building is a, and what we've announced is a forwarder. That's Splunk lingo for the piece of code that takes log data and ships it down into Splunk Enterprise. So yeah, it's much more than just a connector. There's a lot more that goes into it than that. There's a lot of intelligence there, and in particular in a way that doesn't drive a lot of MIPS. And in real time, it meets the SLAs of the specific application solutions they want to build is fairly complex. But, you know, we are we've announced the initial version, the real-time version will beta next week or so, and we'll go GA by the end of Q3. Okay, fantastic. Yeah, we're really excited about the opportunity to work with Splunk and the joint value proposition of the solution. Yeah, I mean, as we've talked about before, so we've been doing some survey work and certainly finding that, you know, I think it was over 60% of the big data practitioners we talked to have already moved some workload from a mainframe or from a data warehouse into Hadoop. So, huge opportunity there and, you know, driven by cost, but also by a lot of the capabilities that you can, and I want you to have that data there and unlock those traditional systems. There's a lot more you can do with it. And one of the things we're seeing just in general around mainframe is the vendors that have specialized in the mainframe for the last, you know, four decades generally have a set of products and revenue streams that are kind of dependent upon that data staying on the mainframe. And so they're not as kind of, they don't have the incentive necessary to collaborate with Splunk or Cloudera or Hortonworks or Vertica to figure out how to move that data off-platform into these environments. That's interesting. We do. Yeah, the data warehouse vendors, you might substitute them for mainframe vendors. Yeah, sure. Similar. Josh, I wanted to interrupt you. We have to go live to the keynote that's going on right now. The CIO from the U.S. Postal Service that we had on earlier. The world's changing. His interview was great. We joked, will you be delivering packages by drones because Amazon's right on nipping at your heels. We really appreciate you coming on. We'll see you in New York City. Josh with Syncsort. Josh Rogers, the president of Syncsort. Doing it all with Cloud, DevOps. Thanks for coming on. Thank you so much. Appreciate it. We'll be right back at the short break. This is the Cube.