 Okay, we're back live inside theCUBE at SiliconANGLE.tv. Exclusive coverage of the H-PACE conference. This is the first conference ever of the H-PACE community's putting on as part of Hadoop. And I'm here with Norbert Berger from TBS. Turner Broadcasting. Broadcasting, TBS, T. Turner. You guys have been doing a lot of big data, I've noticed, and a really high-tech organization. I've bumped into some of your cohorts over the years at events, intimate dinners and conversations. Are you very active in the community's company? You're here at the H-PACE conference. Tell us, Juan, why you're here and what do you think of the conference? So, we at Turner have been using H-PACE and Hadoop for about a year and a half now. We've got a cluster running actually in production that we're using to analyze video operational data. So, metrics about video playback sessions on our websites. I'm here, I think, because number one, I think the conference is a very targeted, very effective way to reach a lot of the folks in the community. And also, because I think at this level, there are a lot of open questions about how H-PACE will mature as a platform, how it will scale, and I think this is a great place to learn about it and to hear from other people. What's your role there? So, I'm a principal architect at Turner, and that means that I'm impossible for choosing the technologies that we use at Turner for answering some of these big data questions. Mentoring other folks in our team, making sure that we're making the right decisions on a technical basis. And specifically, a principal architect in our video team, video, we deal with both the front end and the back end of video playback. And, again, what we're using H-PACE for is to track video player sessions across all of our partner websites and then to understand important metrics from those sessions, how people are gonna performance, how people are seeing from video playback, what kind of errors they're seeing, what kind of results they have, what their sequence is, essentially, the clickstream of data for video on each of our websites. And H-PACE is a good fit for that. Obviously, we were just talking before we came on some of the things that we're doing in media companies. And it seems that all media companies in general have woken up recently to big data. Go back and tell us about Turner. You guys have always been kind of cutting edge, but when did you guys realize, hey, we got to actually think about using this big data to really, when did it hit the radar for you guys, that this has got to be real priority? I think when you think about the number of sessions that, the scale of the sessions that we deal with on a day-to-day basis, we have within Turner kind of a core consulting group and that's responsible for making technical decisions on for each of the properties, not just one property. And when you centralize those technical decisions and you realize the number of visits that we have to all of our websites, the choice of big data becomes pretty clear. It's also, I feel that in this day and age, whether, if you have any kind of presence on the net, you're gonna get kind of pulled into this big data feel whether you like it or not. Just because of the scale of the internet and essentially the reach of the internet. I mean it's not that, any kind of reasonable website today is going to get the kind of numbers that really make up a big data decision, fairly evident, it's not. Well also we were talking a lot of folks here inside the cube, it's great, I love the cube because not only do we get to share the knowledge of you folks out there, but I had the pleasure of interviewing really a lot of the brightest minds here certainly in this vertical it's growing and that is personalization. So one of the things that's interesting is that the notion of recommendations engine we talked with Christophe and Webidata and others is that now the ability to provide some personalization. So you guys have huge audience, so you have a big audience. So narrowing it down and giving someone a really good experience on their interests is probably something on your mind, right? It is, certainly. At the moment we're focused on kind of operational performance of video playback sessions but certainly questions about business intelligence, questions about recommendations, that's not too far off. That's not that. So first year goals, get in production. Right, right. Get in production, understanding again what kind of experience people are having on our websites and making sure that we're, that video player sessions, that there aren't issues with video playback sessions. Then from there, kind of that second order of analysis of understanding trending, understanding recommendations, that's kind of the next phase for us. Let me see what the questions are from the Twitter sphere here. So one question while I look this up is, what are you guys seeing in terms of developers, kinds of programs you guys are hiring at Turner? So our stack actually, from a technology point of view, we're using some of the major components that are being caused around here, certainly Hadoop and HBase. We're using a pig for analysis purposes. We happen to be, my group is actually using Python primarily, so we're mostly a Python job, but certainly Java is a big technology here at Hadoop and HBase, given that the core components are written in Java. So both of those are kind of interesting technologies that we're looking to hire more folks with experience in Python, Java, and certainly some of the- How about your operations organization, you guys spread out now, you guys mostly in Atlanta? Operations, yeah, certainly, certainly in Atlanta, but we have data centers throughout the U.S. Engineering centers? Engineering centers, yeah. As well. Yeah, so we're building actually, we're expanding to the other parts of the U.S., but centralized right now in Atlanta. So we are, again, focusing on understanding video playback, performance problems, errors that people are having with video sessions on our websites. Tell us what's the most provocative conversations and sessions you've heard here at HBase? What are some things that have, wow, that's something I'm going to have to think about. Really compelling. I think some of the sessions, I just came back from a session from Ian from Salesforce. He gave a great talk on schema design. I think that's a really interesting topic that is not well understood by many members of the community. I also feel like some of the talks earlier, I think Cosman gave a great talk from Adobe about how they're using HBase for OLAP style queries. That's really digs deep into some of the best practice of HBase, how you design row keys, how you design column families, how you handle things like nested entities. Those are all, I think, really, really important concepts that really don't have a great rules of thumb yet and slowly through conferences like this. I think we're getting to the point where those become a common language and we're going to be able to take HBase to the next level. So a question from online here I'm reading is, what are you guys developing that's new, that's different from when you get on cable? And mainly around analytics. Obviously analytics in the media business has always been kind of like a crapshoot, gross rating points, but now with set-top boxes you have some instrumentation. So the question is, what's the metrics look like for an audience instrumentation like for you guys? So there's certainly a trend in our industry moving from broadcast to online. That's the direction that everybody's realizing. More and more folks are moving online. Certainly Turner has realized this, we have an effort called, we call TV everywhere, which even some of the companies that you see here are involved in. And it's basically an effort to standardize the video playback experience for both online and broadband and make sure that a paying subscriber, a paying cable subscriber has access to content individually, the metrics I think are very different, but they're going to become standardized eventually. I mean, within the next few years, I think you'll start to see some standardization in terms of how people tag sessions in cable and how they tag sessions online and how they come together and how you represent the same thing. I think still they're very different. They're not quite the same. Who's deciding that? Is it the standards body? Is it more? I think it's really, I think you'll see some of the companies like Nielsen and Omniture probably get together and make some official kind of currency definitions on what it means to have a session in cable and what it means to have a session digitally. So it is certainly something that I think everybody's interested in, but there's not one clear answer yet on it. Norbert Berger with TV Turner Broadcasting. Final question, share with the folks your agenda for the next year or two. I mean, as you go forward, move past video, how are you going to be using HBase and big data for your architecture? Right, so I think HBase is a technology is certainly growing. I mean, I've been telling everybody I meet here today that this conference today reminds me of what Hadoop World was like in 2009. I remember going to that in New York and seeing a conference about the size, about 200, 300 people. And then if you look at Hadoop World today, it's just amazing. I mean, it's grown to thousands of people. I truly feel that HBase is going to grow like that. I mean, I just think that it's just the beginning of the technology. And once we get through some of these technical hurdles, understanding how to deploy HBase and understanding how to host HBase introduction, there's no limit to where HBase will grow. Okay, there's no limit to where HBase will grow. It's going to be a massive industry boom here. Obviously, we're being behind in support of it. We endorse it. I personally think HBase is one of those moments where it's all going to kind of come together and the ecosystem's just about to begin. So appreciate your comments. This is SiliconANGLE.tv's coverage of theCUBE here at HBase Conference in San Francisco. We'll be right back with our next guest after this break.