 Welcome back to Silicon Angle and Wikibonds. This is theCUBE. This is our flagship program. We go out to the events. Extract a signal from the noise. We are here at the Hadoop Summit. The hashtag is Hadoop Summit. Go to Twitter, get on the hashtag and tweet your comments, feedback. Ask us questions. Dave Vellante and myself will answer them. We'll ask the guests those questions. This is theCUBE. We've been here for our fourth season now on the Hadoop Circuit. We've done all the Hadoop summits, all the Hadoop worlds. And we're really looking forward to continuing to bring these two days of live coverage. This is theCUBE. I'm John Furrier, the founder of Silicon Angle. I'm joined by co-hosts. I'm Dave Vellante, wikibond.org. Aaron Davies Morris is here. He is the managing director of Worldwide Professional Services for the Intel Big Data Group. Aaron, welcome to theCUBE. Dave, thanks for having me. John, thanks for having me. It's good to be here. So Aaron, word is that people have been seeing Intel a lot in the field. Yep. You guys made an announcement. We were covering that. We were last strata, I guess it was February. Good. So let me start with why are you guys in the Hadoop distribution business? Well, Intel's in the business for a number of reasons. First, foremost, we've always been a chip manufacturing company and we want to see folks adopt Hadoop broadly. Our distribution, other distributions, et cetera, that helps drive chip sales for Intel and that's a good thing for our shareholders. So that's one reason. We certainly also are looking at, as a company, how do we diversify? How do we grow? How do we find more high-margin businesses? So things like services around Hadoop, selling Hadoop distribution licenses are all good for our business. It's very disruptive in the data center. Obviously, the buzzword is software-defined data centers where people are kicking around. We're going to see a lot of that at VMworld upcoming. We're going to be the cube there. But that's kind of marketing right now. It's really kind of an elusive term. Software-defined data center is one of those terms that's like the moonshot. Yeah, we'll get there someday, but right now you're seeing software-defined networking, software-defined storage and servers. That's infrastructure. Now you have software-defined Hadoop in software. Hadoop is going to be a big part of growth in this new era of managing the tsunami of data. Okay, it's going to push the envelope on performance and obviously reliability. So I got to ask you about things like security and the platform. We had Rob Bearden on the CEO of Hortonworks, and he talks specifically, hey, there's developers out there. We want the program on top of Hadoop, and the Hadoop platform itself is maturing fast. There's still a lot of work to do. So I got to ask you, what areas are you guys working on in Intel to one, make the chips faster, smarter, more reliable, at the same time move the needle on the Hadoop ecosystem? Yep, hey, John, thanks, and I appreciate that. We are, we're committed to open source. So as we develop our Hadoop distribution and we develop things specifically around security and governance, which we're spoken about a lot this morning, and I think are a big, big thing that needs to be addressed. I came over recently from McAfee. My whole career has been in the security space. I was really excited about the opportunity to come over here because of what Intel's doing in security and governance of these solutions. For me, I looked at it like, hey, you can't take a technology like Hadoop where you're pulling all this information in and not put the right security controls on it. So Intel's focused on improving performance, both for folks that use our hardware, our chipsets, and also other technologies, right? But improving that so that you can do good, strong encryption, file level encryption to make sure that the data that's getting pulled in is not exposed if it's breached, is a core thing for us. Another thing we're starting to do now, you're going to hear at the show about Project Rhino. It's a big, big deal for us, enhanced security controls around access controls so you can tell who's getting to what data. That's a core thing for things like PCI and HIPAA. You cannot be PCI compliant if you don't have the right encryption. You can't be PCI compliant if you don't have the right access controls. You also can't be PCI compliant if you don't have the right auditing capabilities. The ability to turn around and say, I know that John got at that data, very critical item from a governance perspective and something that's got to be looked at. And we're really excited about what we see here when you see things like some of our competitors and new startups that are focused in that space. We think that's a testament to the fact that back in February when we came out with our distribution, we made security a core thing for us. And we think that's shown here and that the rest of the market is starting to lean towards it. I felt like that was a big motivation of that announcement. We had Boy Davis on and obviously that came up substantially. And Merv in his talk today said, security's a weak link. Yep. You know, called that out and kind of, we kind of knew that. And you're the McAvie acquisition obviously huge. You guys attacking it, you know, from the core out. You know, I guess pun intended. Talk about what's happening in the field. So you have an organization. Maybe it's some line of business. Shadow IT or something. It says, all right, we're going to go do some big data project marketing. See that we see some potential here to mine some data. Security's kind of an afterthought. They don't even bolt it on. They don't even think about it. Yep. And then somebody knocks on IT's door and says, hey, you mind checking out whether or not they comply? And this is big dissonance. Yep. I presume you see that. How do you see that gap getting closed and what specifically can Intel do to close that gap? So, you know, Intel's, in Merv's talk this morning, he talked about the big dogs and the little dogs. And we're certainly a big dog from an organization perspective. We think our partner strategy helps us meet large enterprise needs. Those organizations, when they start having both planned and focused projects as well as you said, shadow IT projects, creating these data lakes that we're talking about this morning, the objective of those data lakes is to pull all this information in irrespective of what it is, where it's coming from and allowing future folks to make good use out of that. Those data lakes are huge places that bad guys are going to go look for. I spent 10, 12 plus years focused on breaking into organizations and trying to find this holy grail. Intel's focus is how do we protect that? So our partners are going to help us. Companies like Dell and other organizations that we're going to look at and say we can do core technical deployment, right? Those folks are going to help us in terms of doing things like enterprise data management. We'll go to McAfee and other organizations to help us with security. And in doing so, Intel will be able to move the ball forward with helping organizations put in place a real complete Hadoop solution. I think right now to your point, folks coming in and just standing up infrastructure is what we saw years ago and that's going to lead to, I think, a raft of breaches. You look at bad guys getting in, they're going to go after that data right away. Well, and the anatomy of those breaches has changed over the last decade. I mean, it used to be you get in, make a lot of noise, ha, ha, ha, stick a virus out there. Not at all. Now it's stealth. I often quote the stat that after a breach, it's on average 400 and some odd days before you find the perpetrator. So people don't want, the bad guys don't want to get found these days. Not at all. So the investment seems to be shifting from protecting the perimeter, the moat around the castle, to trying to discover people when they've got in and get them out. How do you use, I mean I know the industry is using analytics to do that. Is Intel participating in that space in earnest or is it really just powering the analytics? We're powering it from a professional services perspective. We've got a number of folks on our team that are focused on those sort of things. We have a bunch of data scientists and one of the things that we think we can do is help folks with rapid sort of security assessments of their infrastructure to help in that space. I also agree with you that, bad guys are focused on sitting there very quietly. At McAfee we released a white paper around Operation Night Dragon that foreign governments were sitting there putting people into large utilities. That's public out there, you can go grab that information. And the idea was for them to suck all this core data out, things like well head data so you can move markets, et cetera. And having a layered defense of security is where you got to go. So the things that we're doing around encryption, the things that we're doing around access control and auditing are one set of that layer but you also need other technologies around that. Everything firewalls to IDS to network, DLP, et cetera. And again, Intel's focus is a large organization as an enterprise provider of services and solutions is to turn around and say we can bring those different parties together. Technical partnerships like we're doing with Cray, technical partnerships like we're doing with Dell, services partnerships like we've done with a number of organizations. Well, so what's the nature of those partnerships? They're essentially distributing your distro as part of their solution, is that right? Well, they vary but yeah, most of the time we've got a relationship where they're building our distribution on top of their hardware, right? Or in the case of services providers that we have relationships with, they're providing a number of resources that are going to help us deploy technology or do services that we would not be able to otherwise do. Aaron, I got to ask you the question honestly. We've been covering Hadoop since the ecosystem started within the hub four years ago. The always the question was when do we get to production, POC? And then the analytics market hit really two years ago where I saw analytics as the killer app. Now you see an application tsunami coming, people are building apps. But enterprise customers and large scale dudes are still in that large proof of concepts and or some stuff in production. What are you guys seeing right now in terms of the key critical metrics that customers are evaluating if they do port? So first up, we're seeing POCs that range anywhere from a month to three or four months, sometimes a little bit longer but we've definitely seen a compression in our POC timeframe. And one of the reasons I came over here was to help make that a more kind of automated process and smooth that out so customers can go from POC to production very smoothly, seamlessly and quickly. Okay, Aaron, thanks for coming on theCUBE. Check out Intel, they got a lot of stuff going on there. Their own distribution of Hadoop, obviously they're investing heavily. It's early in the Hadoop cycle, congratulations. And I think security and adding hardened technology in there to abstract the way the complexity has always been an Intel strategy. And obviously bringing it out to large scale deployments would you guys have customers there in the data center and beyond. So congratulations. We're here inside theCUBE with Intel talking distribution security. This is Hadoop Summit. This is theCUBE. I'm John Furrier with Dave Vellante. We'll be right back with our next guest after this short break.