 Live from New York City, it's theCUBE at Big Data NYC 2014. Brought to you by headline sponsor, Juan Disco, with support from EMC, Mark Logic and TerraData. Now, here is your host, Dave Vellante. Welcome back to Big Data NYC everybody. Really pleased to have Sean Conleon from Hortonworks, a longtime guest and Chris Tugut from TerraData. Welcome gentlemen to theCUBE. Thanks for having us. Thanks, good to be here. So, really interesting things going on and you guys are at the heart of it, right? Chris, I remember when TerraData first came out, I was at IDC at the time and you guys came to the IDC, with the big dog and pony show and we're like, wow, this is cool stuff. It's really going to change the way in which people look at data. It was really the first time I personally started to think a lot about data and its business impact and it's been amazing to see the run that TerraData's have. Meanwhile, you've got Hortonworks and all this big data and Hadoop thing going on and the two worlds are colliding in a really interesting way and your two companies have collided or partnered in a really interesting way. So, Sean, let me start with you. Give us the update on what's going on with Hortonworks and your perspective on the relationship with TerraData. Sure, yeah, I mean, it's funny. I just passed my third year anniversary at Hortonworks so we're just a little over three years old. I joined shortly after the founding. Oh, man. Three times seven is 21 years. Yeah, exactly. Well, I don't know what Elfin years are. But since our founding, it was really, how does Hadoop fit into a broader data architecture? And early on, really out of the gate, we formed a partnership with TerraData. I think a bit of it is the roots on the deep engineering optimization work that you guys have done in your systems and how do we take that and extend the data under management so that you can get a more sort of unified experience for customers. Fast forward to today and we kind of continue to execute together in appliances, software, deployments, et cetera, and customer sites. Like I said, the three years have gone by pretty quickly. The, where the market is right now is it's accelerating. I think even particularly beyond the Stratahadoop where all the other conferences I think Hadoop is really firmly on a lot of Seattle's radar screens. It's pretty exciting. So Chris, when you're as successful as a company like TerraData is, everybody wants a piece of your hide. And so you got big customer base, a lot of success. But the difference between my observation between companies today and the ones, let's say the mini computer guys, I'm from the East Coast and I saw denial. Ken Olson, Unix is Snake Oil, An Wang, LPCs who needs those things. Companies today embrace the change. They see it coming. They're not in denial. TerraData is a good example of that. You guys have made acquisitions. You may have strong partnerships. What's the conversation like sort of strategically inside of TerraData and how are you bringing that to the marketplace and your customers? Yeah, I think it's a great comment, right? You can't stand still in this marketplace and clearly with our acquisitions of Think Big and being able to have consulting services around open source and being able to get value from data. And then our acquisition of RevLytics, very specifically around Loom, to be able to do integrated metadata, data lineage and data wrangling, all in a self-service UI. And then also our acquisition of Hedapp, really a pioneer in doing SQL on top of Hadoop. And so I think all of this is about helping to extend into a broader analytical ecosystem. Because it's not one system can do it all. There's a lot of very unique analytic engines out there. File types, economics in terms of how people wanna store and analyze data and really the vision that TerraData has is we wanna help orchestrate that. Help orchestrate it all together in a way that makes it transparent for the user but enables them to really take advantage of all the power of all those different systems out there to get the best insight against the broadest set of data. Well, it's interesting when we survey our customer base and ask them what kind of tools they're using and their big data deployments, the number one and number two is data integration in their existing enterprise data warehouse. So a lot of people ask us, oh, is my data warehouse going away? Maybe 100 million years from now. But so it's fundamental, at the same time, customers are trying to figure out, okay, where do I place my bets? Where do I put the investment? So Sean, what are you seeing and what are you advising customers? Whether it's R&D stuff on Hadoop, whether it's real production, what are you seeing out there and what are you advising? So the successful deployments, and this is really accelerated not only through 2013 but through 2014, is those folks who are focused on the new forms of data and maybe blending it with some of their existing forms of data for either enriched existing analytics apps or net new apps around sensor data or a lot of these new forms of data. They're the ones who, from a Hadoop embracement perspective, have sort of repeatable momentum and it moves very quickly, right? So while I, it's always interesting to talk about the depth of this, the depth of that. What customer wants is they wanna optimize architecture that gets the data into the right system for the right application and right sort of delivery to the end user. And they also want to figure out these new types of data they're flowing in. How do I optimize my supply chain, right? You might have vehicles with sensors that are part of that supply chain or pharmaceuticals that are delivered to the drug stores and are you getting sort of the right meds into the right locations to head off outbreaks which we're, you know, assembling just very real and dynamic. So it's optimizing supply chain, optimizing sort of around the customer. Single view of product, those types of things. There's new forms of data and existing data that they're trying to figure out how they bring that together to unlock new apps that they weren't able to do before. And from our founding, we were like, this is all about how do you maximize your data under management so you can actually achieve that goal, right? And that comes from an integrated architecture, right? At the end of the day. It isn't one system or one product. It's actually an architectural approach. Chris, I wonder if we could add to that. Customers tell us, you know, data warehouse environment's a challenge. We always, anytime Intel comes out with the faster chip, we got to throw it at the problem. Data ingestion is just amazingly unpredictable. Our budgets aren't going up, but the data is rising like crazy. So how are you helping customers sort of address that problem? And it relates obviously to what you're doing with Hortonworks and other parts of your ecosystem, but I wonder if you could talk about that directly. Yeah, that's a great comment. I mean, if you look at 16, 17 years ago, Teradata really embraced this idea about an enterprise data warehouse, right? You bring your transaction data together and you manage it in a way that you have that consistent, shareable, govern data across the enterprise. And then this onslaught of what we all talk about, big data, right? New data and new data types. It's very funny because I hear people say, well, aren't you concerned because all the sensor data and the web data, you know, going in Hadoop and that's going to take it off of Teradata? Well, the reality is Teradata always had data that was there that was available for high value workloads in which lots of users needed consistent concurrency against its environment. So for us, and it's exactly what Sean said, it's all about bringing all of this together in a scalable way. And we do it through our unified data architecture. And our unified data architecture includes Hadoop and our partnership with Hortonworks. It includes Aster and our solution for being able to do discovery and discovery analytics and Teradata for being able to share data in a very govern, cleanse way across the enterprise. And then one of the things that we do with the unified data architecture is we glue it together through Teradata Query Grid. And we pioneered this with joint engineering with Hortonworks. Originally it was going against H Catalog using SQL H and we've expanded that now which is our query grid platform which can push down processing. So if you've got a lot of sensor data and you wanna do some aggregation and process that inside of Hortonworks and then take some of those results and combine it back up in the warehouse where I have all my warranty information, where I have my customer service information, where I have people hitting it consistently all the time. It's a great marriage and it helps companies scale in an economic way, yet get the value across the unified data architecture. Actually accelerates the adoption as well because it actually provides a more of a unifying way for in a consistent way to get at that data. I wanna talk about partnerships because in an exploding market like this where you've got a huge number of vendors coming out, you've got some guys at the core, everybody wants to do partnerships and John Furrier calls some of the partnerships, he calls them Barney partnerships and others have substance. How do we parse through that? How do we understand which ones have substance? I've heard joint engineering, joint go to market. Can you talk about what makes a substantive partnership that's gonna deliver ROI versus one that's sort of, yeah, let's do a press release. I wonder if we could maybe talk about that a little bit. Sure, I mean, I think in the early days of Hadoop, a lot of it was around connectors that would move data from systems and that kind of stuff, not a very optimized experience, right? When you're talking about things like query grid that are intelligent about, all right, we know this data's in Hadoop and I can push it down and get the benefit of the bargain of leveraging Hadoop for what it does and pull it back in a familiar way. That's a much more sophisticated technological experience that actually requires engineers from Teradata, engineers from Hortonworks to actually roll up their sleeves, create multi-year roadmaps and start delivering on this stuff that doesn't happen overnight, right? And so when we talk about sort of deep engineering work, it's what's next to come, right? How do we begin to apply some of the principles that Teradata has had from workload management and hot, warm, cold data management and extend those same principles into the broader sort of data fabric that Hadoop exists in. They're really interesting engineering conversations that can happen and that type of partnership that we have provides the vehicle for doing that. So Chris, I wonder if you can add to that. I mean, what makes this relationship substantive? I mean, convince me, let's say I'm a skeptic, oh yeah, just another deal. Why is the Teradata Hortonworks partnership different? And I think it's absolutely what Sean said, but I think it even goes beyond, right? It's about engineering, it's about go-to-market models, it's about sales alignment, it's about joint marketing as we're bringing it to the environment. But I think one of the biggest keys is the engineering. It's not just doing it once, it's continually looking at what's going on in the marketplace and looking about how we can do joint engineering to bring products to market. And some of the examples of this certainly is query-grade. There's other examples where we've built in Viewpoint where Teradata can use a single plane of glass to look at both the Teradata data warehouse, the Aster discovery platform, and Hortonworks all in a single environment to make it easy for users to manage that complete foundation. We worked with Teradata Vital Infrastructure to add Hortonworks into our appliance so that we can deliver a platform to our customers that's ready to run up and running right away and one-stop shopping. So if a customer wants to come to Teradata and wants to say, hey, I wanna get all the support for the complete UDA, then we can do that and we work with our partners, Hortonworks on the back end to make sure we have that linkage for level four and five. So I think it's that tight integration throughout the supply chain of what you're doing and how you're going to market that really makes a significant difference. And whether they wanna consume it in integrated appliance with everything or software or subscription, being able to sort of engage through Teradata for that relationship where they provide that face to the customer in a familiar way I think is important. And that is one of the things I'd point out. Three years at this at Hortonworks, that was a founding principle was how do we partner with the data center players in a way that the end customer who trust that vendor can procure this new capability in a way that's very familiar and sort of de-risks it from their perspective. I think that's a great point. And last week, together we announced the Teradata Cloud for Hadoop. So it's not just an appliance model. If people wanna be able to go up into the cloud and test Hadoop, bring Hadoop up in a cluster, they don't wanna have to worry about getting all the data scientists and doing the managed services. Teradata's put an offer together where they can do that. They can put Hadoop in the cloud, they can have Teradata in the cloud, they can do hybrids of it. So we're working together to bring multiple deployment models to the market. Cloud meets big data. So you guys have been down at the show at Strata and Hadoop World. Other announcements? Anything else that you guys announced? Yeah, we announced Teradata Loom, 2.3, this was from our acquisition of Revolitix. And like I said, it's a key foundation, especially as we start to see more companies looking at deploying Hadoop in a data lake type of model. Because the more you have different disparate data types in there, you need to understand lineage, you need to understand governance, you need to be able to do wrangling and bring it together so that that data is accessible to the broader enterprise throughout the UDA. So we're excited about that product that we brought to market. And you guys announcing anything here? Yeah, so yesterday we announced that we're working with Data Platform 2.2. A lot of new capabilities in there, not the least of which is things like HBase, Cumulo, Storm, and others, and Spark are on yarn. So they're all integrated in, delivered easily. And I suspect you'll be talking to our Microsoft counterparts about capability out of the box when backing up to Azure Blob Store from one-prem clusters and automating that process. There's a lot of cloud goodness as well as capabilities that are in the platform, but a very significant release. A lot of new tech. You mentioned Cumulo two years ago. The two Hadoop summits ago, it wasn't quite two years ago, but two Hadoop summits ago, security really started to get front page in this world. Before that, nobody was paying attention to it. It was a signal to us that, hey, this is getting real. Yes, well, and even like it's interesting to see the dynamic, so as you provide both Apache HBase and Apache Cumulo, and the HBase community has added a bunch of security capabilities in there. So it's really interesting enablement of choice. And then our acquisition of XA Secure earlier in the year is manifesting itself as Apache Ranger. So that's all gone, submitted to Apache. Centralized authorization. Again, security is paramount for adoption. Thanks for the contributions. All right, we gotta leave it there. We gotta go. Gentlemen, thanks very much for coming to theCUBE. It's great to see you. Thanks, Dave. Thanks for your time. All right, keep it right there, everybody moving back. This is theCUBE, we're live at Big Data NYC. We're right back.