 Okay, we're back here live in Silicon Valley. I'm John Furrier, founder of SiliconANGLE.com. This is theCUBE, our flagship telecast. We go out to the events. We're here at the Cassandra Summit 2012, and this is where all the action's happening in Big Data today. And we're on the ground covering it. My co-host today for this event is Jeff Kelly from wikibon.org. Jeff is the leading analyst in Big Data. Jeff, what do you think? I mean, we've had some guests on. We had a nice run. What's your take? If you're a noob, you shouldn't be here. Well, what I learned is, obviously theCUBE is, the greatness of theCUBE is well received in the Cassandra community. Thanks to DataStacks for really pushing us to get here. And we believe in this. So we're here editorially tier one event in Big Data, DataStacks worth covering. Putting that aside though, I really was out to understand Cassandra's use case. It's clear that it's about scaling. It's about transactions and being bulletproof. And the other thing I was trying to find out, Jeff, was the personality of this community. And I guess I'll just say it's like hardcore geeks, alpha geeks, guys who pride themselves in writing clean code, kind of an elite force. Kind of like things we've experienced with Ploudera and Billy Bosworth, the CEO. Obviously friends with Mike Olson. They're friendly. They work compete together, work together. But it's to Mike Olson and Billy Bosworth to the world that create this kind of culture of excellence. And I'm really, really impressed by the, I don't want to say star power, but the IQ power of these communities because it's a complicated environment. It's growing very rapidly. And it's crossing the chasm. So the good news from my standpoint that I'm happy to report is the community is vibrant. The solutions are solid. They know what they want to do. They want to make it easier to use. So they're on the same path and trajectory as Hadoop. So that was a check off. To me, the big surprise though, was an ironic connection with what we're doing. And that's converged infrastructure. Solid State has so game-changing. And obviously we're covering on Silicon Angle, Wikibon, the work you guys are doing, and David Floyd in particular, is that Solid State is more of an enabler than I even thought in this area. And that is that Cassandra is very well differentiated with Solid State in such way that people are saying that Cassandra is the ideal solution. Cassandra File System, Cassandra Database, for Solid State. That's a surprise. Transactional big data environments and how important that is. The idea that you can just kind of slap on a traditional business intelligence tool on top of Cassandra or any other, no SQL store or big data platform, it really isn't how it works. So it's really interesting to hear companies like Jazzmer, Solid, like Actuate, and others talk about how they're reporting against multiple data sources. Bringing analytics to big data repositories from multiple sources. We talked about, in the past at Wikibon, we had a great pair of insight with Atibio and Relay Technology a couple weeks ago. And they talked about bringing together unstructured data, in that case, to support analytics around pharmaceutical investing. But the whole idea that you need to find new ways to bring together different types of content and different sources. And we're seeing partnerships here between data stacks and Jazzmersoft and others to do just that. So to me, that was an interesting part of the day. Just some key themes that I'm seeing out here, just looking at my notes. I'm sure this'll be, you know, obviously typical in SiliconANGLE fashion. We're uncovering the signal from the noise. You'll soon see this on most of the tech blogs soon. You'll see headlines like this. So these are the themes that we're uncovering soon to be on other tech blogs and other environments. Obviously the relational database versus the NoSQL movement. People are trying to pit that against each other as a war. Totally not the case. Schema is good. Schema less, or less schema is where most of the action is relative to unstructured data. But ultimately it's not about one of the other that's obviously exclusive. That theme is very resonating very, very strongly. NoSQL is the real deal. It's moving into a mainstream audience and you can see more structured database and relational databases around that in work in conjunction with that. We're already seeing it with Cassandra. They're working with Hadoop. Even amongst themselves and NoSQL, you've seen them work with each other. So the other one's tooling. The UIs, the GUIs, the frameworks are being built out to build on the robustness of Cassandra. That's a theme here. And then the specialism versus general purpose. We're going to see that continue to evolve around skill sets, use cases. Just from a general market perspective, you're going to see that where you're going to move from, oh I need to hire that special guy to more of a general purpose marketplace product. And that's where the innovation's going to be. I think whoever can make that happen at all levels will be a winner. That's classic innovation. Abstract away complexities and you're a winner. The other theme is the impact of SSDs and the multiple data centers. This specifically highlights our research and our work on covering the notion of hybrid cloud is where the action's at. Not private, not public exclusively. It's in the middle, on-premise and off-prem working together, that interoperability. I think the Cassandra theme here around multiple data centers is critical. And in fact, one of my favorite soundbites this morning, one of them was SSD is close to a silver bullet as you're ever going to see. That came from Jonathan Ellis, classic soundbite. And finally the theme here is this is all about developers, developers, developers. We are living in a renaissance of developers again. A continuing cycle of innovation. I'm excited to see that. And this comes back down to the specialism. The developer communities are now growing up. Open source communities now grown up. The younger developers are coming in. A lot of mentorship. We heard Adrian at Netflix talk about the evolution of building on Amazon. He's doing a demo standing up 24 node cluster and firing the chaos monkey at it. This is where it's at. People are going to continue to build their business on Amazon. And then instead of having a fail on rearchitect, you're going to see best practices around these kinds of bulletproof databases. So these are the themes. Developers, entrepreneurship is a port of it. So those are my core nodes so far. And we've got some more interviews to do. You know, I just, just a little versus general. I mean, I think that's really, really dead on. You know, the one thing from a technology perspective that we see data stacks do, they're in a fresh platform. Brings together, not just the same, but you can solar. The idea is you're not going to, you don't want to deploy three clusters. You do three different jobs when you can deploy one cluster and do multiple types of jobs on that one system. So that's, you know, Cassandra, data stack has taken that approach by bringing in the three of those together. And then we're seeing, you know, similar approaches with work works in there and the yarn projects, which is essentially next generation map produced, applying new types of processing techniques, besides mapping. So you can do some real-time analytics. You can do some graph analytics and different types of workloads within the environment. So, you know, I think you're absolutely right. And I think we need to see these applications that should take these big data platforms continue to evolve rather than focus on just one special use case. Well, I think it's like the way I would look at an operating system, right? Just subsystems within an overall operating environment. And I think, you know, brought up Hortonworks is a great example. You know, it was Cloudera was the only horse in the race in big data, in comes Hortonworks. But what's interesting about the marketplace that we're seeing here in the conversations about Cassandra is clear that the Hortonworks business model is actually set up pretty nicely because what you got Billy talking about and some of the key folks at Datastax talking about is that I want to bring the Cassandra file system in. I want to sit and work on top of Hadoop. I don't want to have to lock out a Hadoop implementation. I want to manage the data and still work through Hadoop. That actually talks about some of the benefits that Hortonworks doing and kind of puts Cloudera directly frontal against Cassandra. So Datastax and Cloudera, where Hortonworks is kind of like this Switzerland. So interesting dynamics. And you know, Pat Gelsinger said on the queue when he was still at EMC that he thinks that there can't be a red hat of big data of Hadoop and that that's just a different dynamics in this marketplace. So again, that's something that we're following. And what do you think about that? I mean, you know, does that change the landscape? Is Cloudera going up against Datastax? Or is Datastax going up against Cloudera? This is how you word it. Cloudera is obviously doing well right now. Financially, their heads down. They're, you know, I talked to the folks over there. They're constantly busy. I just saw Michael Olson and some of the folks over there. And again, they're cranking out. They're so busy they don't even have time to lift their head up. To run multiple workloads in your Hadoop environment. Whereas Hadoop, Hadoop can't just become another silo because if it does, it defeats the whole purpose of big data. It's bringing multiple sources of large volumes of data together. So that's why I like what Datastax is doing with their approach. I like what Hortonworks is doing with their approach bringing together multiple areas, multiple data sources, internally to Hadoop rather than kind of creating silo and deployments to SQL here, Hadoop here, it's on there. So, you know. You mentioned Hortonworks. I talked about Cloudera. This is a startup ecosystem developing. Squirrel, SQLL, I guess it's called. You mentioned Hadoop. Talk about Hadoop. What do you like about Hadoop? It speaks to the same issue. So Hadoop is working on integrating native SQL into Hadoop environment. So the idea is you don't have to move data between your MPV data warehouse and your Hadoop cluster. You can do it all in one platform. So in theory, I really think that is a winning approach. But from a technical perspective, it's obviously challenging. I mean there's a reason people, no one else has built that kind of platform yet. So from my perspective, I think we will get to that point where there is a more standardized all-purpose big data platform. We're seeing different companies take different approaches to it, adapt, takes their approach, data stacks is theirs, Hortonworks and others. What'd you think of Amazon's presentation from Adrian from Netflix? Obviously, you know, Cloud providers, he came on record and was very candid. Great interview, tech athlete, the ideal Cube guest. We love guys like him. Very cool guy, he can handle himself. But he did admit that they do only have one Cloud provider, that's Amazon. Do you see that as a risk for big data providers? Do you think that there should be some fault tolerance, some high availability with other Cloud services? So we're separating West Coast. So you can do that with one provider. Should you put all your eggs in one basket, so to speak? I mean we're getting to the point where Amazon was kind of the only game in town really, kind of really made this whole idea of a public cloud private and real, to the extent that we see that ecosystem grow, there will be an opportunity to spread the risk, so to speak. The thing with Amazon is that whenever they do have a glitch if you walk off at some downtime, it's broadcast, it's on the front page of the New York Times. But overall, it's a very stable environment. So I think to some degree that's a little bit overblown, but it's worth thinking about in terms of your backup. Well we got a next guest coming in who's the co-founder of DataStacks. We're going to sit down with Matt. My take on this is simply this. I think Cassandra is an ecosystem that's dynamic. It's under-hyped relative to other environments. It's a set of alpha geeks with the personalities. They're DevOps guys. They're a lot of more DevOps than I thought. Clearly more scalability mindset. When I used to be a coder, it was like the purists, guys who can get down and write the shortest number of lines of code. They really take pride in their coding. They're mostly, a lot of DevOps guys are doing cutting-edge work. At the same time, they have to deliver on production. One of the criticisms we heard about Cassandra that I think that they're really addressed over the past year from what we've been hearing here in the hallways is they've been getting dinged in the production environment, but there's a lot of disconnects around the data that we've seen. I've captured it in that area and that is that the documentation has been poor as noted by their own people and that's getting better, but it's hard to work with. So it's kind of like age-based other environments. It's really difficult to work with. Mongo has had great success because of its usability, but I think in this environment, Cassandra got dinged. Now what I'm hearing is that they're going to get better than that. The other take is I think they really have an opportunity there, again, under-hyped. What I would recommend for data stacks and other people in the Cassandra community is rally together, better market yourselves. The community is very, very relevant, but the marketing is just, I don't know what you do about that. I think it's more of a solidarity issue. Obviously, there's no frowns. No one's really getting down here. People got a spring to their step, Jeff. I've been looking at the hallways. It's cool. I mean, Cassandra's a cool environment, but again, they got some, if they're the NASCAR car, they got a slingshot on the next turn and move out in front because Hadoop is just owning the hype right now and Mongo is owning that as well. So again, Billy said, Billy Bosworth said, I hate the word big data. I mean, that should be a blog post, but again, that's my take and I think it's in good shape. I wouldn't disagree with you at all. I think that's one of the challenges when you've got a passionate community around technology, not always kind of the marketing aspect doesn't necessarily come natural. It's people who play in this environment. Well, I gotta get above the noise. Okay, that's our quick editorial take on so far. We'll have more guests coming. After this short break, we'll be right back with more commentary live from Cassandra 2012. I'm John Furrier with Jeff Kelly of Wikibon and SiliconANGLE. We'll be right back.