 All right, so I want to come back to infrastructure. You're talking about moving beyond infrastructure, but there's still a lot of action going on there. Cloud Air, MapR, Hortonworks. You have Adapt now entering the scene. What's going on there? Who's winning? Who's losing? What are the key factors? Well, yeah, to my earlier point, it's not to say that infrastructure is not important anymore. There's still a lot of development to go happening and a lot more improvements need to happen to make Hadoop and other, early Hadoop, quote, unquote, enterprise ready. So what's happening there is, the way I look at that market is, you've got kind of the three big players that have been in the market now for a year or more. Cloud Air, Hortonworks, and MapR. And the way I look at it is, MapR has taken the more proprietary approach they're delivering. Their premise was that the open source world is just moving too slow. We want to add some of our own IP, harden up Hadoop, and get it to market as quickly as we can. That's heresy, but they're kicking ass, aren't they? They're doing some great business. They're growing and they've got some great customer-proof points. Then you kind of got Cloud Air, which is taking a little bit of a hybrid approach. They've got some proprietary management software around Hadoop, the open source core. And they've been in the business the longest, so they've got that lead, which has helped them become the biggest of the three. But they're kind of playing the middle game, and then I looked at somebody like Hortonworks is really playing the long game. Completely open source, not charging for their product at all, giving it away, seeding the market at this point essentially. And they're going to build revenue off their services offering, but that takes time. And they've raised a lot of money, and they're not in a rush to start really building revenue. And that's Rob Bearden's specialty, right? Absolutely, he's done this before. So the idea is let's seed the market with our product, let's get people using it, let's get people excited about it, and then as they start to go into production in six months, a year, two years from now, we'll be there to provide the services, and that's how we'll monetize our business. That's the long game. So we've kind of got MapR with kind of the short game, Cloudera in the middle, and then Hortonworks is the long game. And then you've got little startups like Hadaptu and some really interesting things. Hadapt is very exciting. I mean, we saw those guys, they won the Best in Show at Strata in the fall. My sources indicate that they've actually got some serious customer traction. I don't know if you've heard that. Now Impala is Cloudera's announcement. Essentially what these products do, if I understand, is they bring SQL and NoSQL together, bringing real time to big data and Hadoop, essentially, and Hadapt, we had Ming Sheng Hong on theCUBE a while back, and he was really taking us through some of the functions that they've developed at Hadapt, things like the ability to do user-defined queries, which are sort of new in this whole world. So it seemed to me, Jeff, that they had a little bit of a technical lead, even though they maybe didn't have a market presence lead. Now Impala is the sort of, you know, Cloudera's version of that, you know, people are, you know, the knock on his arm, he was rushed to market. Cloudera says, no, what's your take on all this? Help us squint through all the fud. Sure. Well, I think, you know, to understand the fundamental idea here, you know, Hadoop is fundamentally a batch, batch and load system. It's not ideal for running what you might call fast, near real-time analytic queries at hot queries. It's a good place to run really deep, complex analytics against large volumes of data, but that takes time. MapReduce takes time to run, depending on the size of the query and some other criteria. So the idea has been, thus far, for most practitioners, early adopters of Hadoop is, let's get, almost using Hadoop as an ETL layer, let's get all our data in there, let's put a little bit of structure around it, and we want to do the kind of fast queries that we're comfortable with, doing it in technology like Vertica or Green Plum or Teradata Aster will move the data out of Hadoop via connectors into one of those databases, and then that's where we'll do our analytics. The idea is, the idea is clearly, she was not particularly interested in that talking point, but nevertheless, the idea with what bringing the SQL aspect to Hadoop is, well, there's no reason to move your data out of Hadoop anymore. You're essentially bringing the SQL database into the Hadoop environment and running SQL directly on top of Hadoop. So that's what Hadaap's doing, kind of from a native framework, they've started their whole company, the technology is from the ground up built to do this. Cloudera with Impala is kind of building this as an add-on. This is the latest development on their platform. So it was- Was the timing of Impala designed to try to freeze Hadaap? Was that sort of a ploy, or was it just sort of two companies on a parallel universe? Well, yeah, I don't know if it was a ploy. I think it may serve that purpose to a degree, but this is where- So Impala's real in your view. I mean, it's getting traction in the marketplace, getting customers. Well, I didn't quite go that far. It's still very early for Impala. As far as, the latest I heard is still in private beta. It's very early days. I've done a little bit of research on the number of contributors to the, because it's an open source project, Clutter Open Source, the Impala project, and there aren't too many, just a handful of most Clutter employees. We should try to find the Hadaap guys today. I heard they're gonna be here, although I haven't seen any of them. I heard they're in production, so let's try to confirm that. Yeah, so it's very early days in terms of Impala, but I don't think it was necessarily time to impact Hadaap. I think it's just a natural evolution of Hadoop. I think we've done some research and we've talked about our fundamental thesis that we believe a unified big data platform or data infrastructure is the right way to go. The idea of connectors and moving data back and forth really is just a stop cap measure for now. So I think to a degree that the market's listening to some of our insights, and to me it was just the next kind of evolution, and I'm not surprised to see it, and I would expect similar moves from Hortonworks from others that are playing in the Hadoop space. You think we'll see an acquisition this year of one of those guys that we've just been talking about, one of the infrastructure players, a big move to try to shake things up a little bit by one of the whales who really wants to play in this game? Yeah, I think we will actually. Where do you think it's gonna go? My inclination, going back to my analysis of the three major players, I think the least likely is Hortonworks, they're very early in their cycle. Kladera's the most mature in terms of being around for the longest, but MAPR is really got a lot of momentum. So I would say it's either gonna be Kladera or MAPR, the most likely candidates for an acquisition. Kladera's talking IPO, I mean, I don't know how real that is, but. You know, it's certainly possible. I think in terms of potential companies that might want to do the acquiring, I mean, they're the ones you expect, I think IBM and Oracle are probably top of that list. Microsoft, not so much. Why not Microsoft? I think they seem to be taking a little bit more of an open ecosystem partnering approach and a little bit slower approach. They've done their partnership with Hortonworks for the HD Insights platform. So Hortonworks might be the natural company they might look to if they're looking for an acquisition, but as I said, I think it's a little early in the Hortonworks cycle, so I don't really see that happening. But you know, Kladera, take it for what it is, but you know, Mike Olson's an Oracle guy in his previous job and, you know. How about Splunk? Let's talk about Splunk a little bit. Big IPO last year, really doing well. Rabbid fan base, users love Splunk. What's the update on Splunk? Splunk, well, so as you may have heard, they are the center of rumors around acquisition as well. And again, the same players are IBM and Oracle or the two suitors. That news was kind of hit last week, it's gone a little dark, we're not sure where that stands exactly, but they're happy. Where they smoke this fire, they say. Yeah, well they had a great year last year. They're doing very well this year. I don't believe they reported full earnings for the year, but they're probably gonna top $180 million in revenue. They've got great momentum. They're now evolving their, you know, Splunk really started focusing on making sense of infrastructure, IT infrastructure data, machine generated data. And they did that with a series of applications, essentially for specific functions in the data center. What they're doing now is kind of opening up their platform really to outside developers. So they've announced several SDKs over the last, in the fall. I recently spoke with a company called Pre-Lert, which has developed a non-lead detection software that they have built specifically a version, specifically for the Splunk platform. So it's gonna run on top of Splunk. So Splunk's really getting now into the platform game, not just, so they're offering kind of their own applications on top of their platform, but now they're also opening up to developers, sort of like the Apple iPhone model. You've got Apple applications, you can run your iPhone, but it's also open up to outside developers and you can run those as well. So I think it's a good move on the whole. The risk for Splunk in that case is of course, if someone comes along and starts creating applications that kind of cannibalize their existing business. But on the whole, I think it's probably a smart move and they're under a lot of pressure from the company to deliver, so. Now Splunk's one of those companies, so many companies out there now looking at it and say, wow, why didn't we think of that? So that's great, you know, they gotta love the innovation. All right, Jeff, we gotta roll, get more guests. Thanks for coming on.