 That's Williams and Clint Finley guys welcome back into the cube from yesterday. You had a good day. What's your how was your day like? I'm really excited you guys to make this trip and we got a great team here today So what do you guys see? You're out there scouring the landscape here. What's the stories there? Yeah, yeah Well, well, we're just putting together just kind of our analysis of the event. We're pinpointing five trends that we're really seeing here you know one is the growth of Hadoop, but the concerns about the ecosystem and Perhaps there's you know some concern about forking or some kind of mutation Training ease of use is another trend that we're definitely seeing dev ops. What would you add to that Clint? Real-time people are really interested in how to do real-time with with Hadoop using H streaming or H base And and and definitely, you know the whole question about the you know the open-source community and how that and how that is developing as As we saw in Doug cuttings keynote this morning. So we're out there You guys are out we're obviously covering the the industry in general the tech industry Services angles doing great. You guys are right some great content up there We saw our news hitting Silicon angles site today a lot of great stories. What what do you guys see? This show and Hadoop in particular impacting the rest of the industry any observations you have We'd like to share as you talk to folks in the hallway here Yeah, I I I'll I spoke to a few people on the hallway today in particular. I had an interesting conversation with a company called travel Travel sailing and another one a conversation with the travel travel velocity and the travel velocity developer said that they are looking at H base to do better real-time analysis on Advertising to get to optimize both their their marketing and their personalization Yet the other fellow was talking about how they are starting to build better measurements inside Big online retailers such as toys are us with to do so they can tweak their algorithms more so they can get to really help their clients understand their their marketing dollars And so those are two examples I think also what we're seeing is how big data is being used for real-time analysis to add features to products They have features to to applications to do that kind of that real-time development Clint, you know, we had a comment here on the Cube from the guy from Cisco Jason and he said you know Jacob and he said this is a great show a lot of technical people be a very engaged community So you know the sessions are well attended great questions What are you seeing from a tech crowd? What are the hot trends here inside the show from a tech perspective? Well, I've been seeing dev ops happening right before my eyes basically with developers and hops guys just Having lunch together from sometimes from different companies and talking about just swapping stories about how to make Hadoop work better and How to make dev ops happen and educating each other So it's been really exciting to see that Hadoop really is necessitating dev ops It's making dev ops something that isn't optional because if you're going to do Hadoop You have to have your developers and your and your operations team like working really closely together You've got to have your operations team doing agile deployments using automated tools and stuff like that DevOps could be what transforms it really could I mean that it's just Really need a whole new set of talents to be able to to monitor these multi-tenant Infrastructure talk about what you mean by dev ops right some people may not be familiar with the term So maybe you could describe it. How would you define it? There's a few aspects to it one of the quick ways to explain it is it's applying the agile methodology from development to Operations that what that ends up looking like is that you a have your your developers and your operations people working much More closely together instead of being completely siloed that doesn't mean that your system administrators are writing code But that they have a really good idea of what the developers are doing and vice versa The developers need to have a better idea of what the impact of pushing out new code is and making changes are so that The said operations can get new changes out quickly It also involves a lot of using new tools that enable system administrators and operations people to work More like developers work by using automated Automation by using scripting tools Doing things that let them focus more on creatively solving problems instead of putting out fires It's a it's a cultural movement too without a doubt the people who are doing it are part of they feel like they're part of an independent Culture inside it so it's like a programmable infrastructure in a sense And Doug cutting said today that when he's not coding he feels like he's not being productive and hence your cultural Angle here. I think it goes back to to what we heard yesterday from Jeff hammerbock or when he was talking about data scientists And how data scientists are kind of like Jim rats their data rats, right? I mean you just you just get immersed in it I think that's part of the dev ops culture well face But we had Jonathan gray on as well and what he was saying was at Facebook There is no distinction between ops and engineering and that you know engineering is ops and ops is engineering and yeah Absolutely, there's no silence. It's like they are working together Clearly and and and you know Doug Gourlet Dave just mentioned on the cube that you know Configuring networking gear has been an ops thing. You know, it's been manual and Now with cloud and you want to provision, you know infrastructure and you can actually do it automatically So I think automation is a big part of that as well and virtualization because now you can actually have virtual network switches And you and of course you have for years now. We've had virtual Servers as well. If you think about like our coverage of the HP moonshot, right? I mean that's a machine now that has 280 servers on our blade And you can have thousands of servers in this boss energy efficient The arm chip was designed for off then on versus just on then turn off So what you what you're going to see is the ability to manage course, right like resources, which is a programmatic issue, right? So now you have the notion that hey, I can spin up Clusters to do that now. I just talking to Eli Collins he had cloud error earlier on we did a segment. I saw your Skype about big big top Do you just did you skype me on? I was it? Yeah, I was just asking about big top Yeah, so big top he came in actually did a quick segment describing an open source project And he was also informing me that were which is Tom white's open source project is exactly That kind of similar DevOps where you could actually program on EC2 to do it's all about spinning up those those clusters So so were is that particular project is going well big top is more of a stack testing and integration So I agree. I mean I think Hadoop is implicitly saying opps as standard a mandate not even an option So you guys have been focusing on adoption at this event you're talking to a lot of the Consumers of Hadoop We've been talking John for over a year now about the gap between cloud service providers and the traditional IT shops And you know the latter the traditional IT shops are doing more with less dealing with budget cuts in 2008 Some of them there were 50 to 70 percent budget cuts. Meanwhile, you got cloud service providers They're investing there. They're they're they're automating and there's this widening gap and my question to you guys is Do you think that? That will continue We can the traditional enterprises keep up in the big data context Or are we going to see that that is a barrier to adoption that lack of funding? I don't think so I don't think they can keep up. I really don't I'm I'm more convinced of it than ever Yeah, you know when we talk about the internet companies in particular, they're engineering They're engineering-centric organizations and this whole concept of developing on the fly like developing Internal tools to be able to manage all this data and then taking those tools and open sourcing them for the community And then that building into more tools, you know and more development that is going to be a very tough model to be Yeah, so so I guess I guess from that we could predict You know the ascendancy obviously of the cloud service providers and people who sell to them and the impact that that's going to have in terms Of disrupting the existing order. I'll just let me add one more thing I mean just look at this post I wrote this morning about SAP's new online mobile app store Here's what you need to do you go to their mobile app store you get a demo of the app Okay, then you get then you get then you contact a sales rep or an or someone inside sales Then they do a consulting Work with you to determine what integrations that you need you follow that up then With the app that you get six weeks later Okay, that's gotta be a totally different model now Of course these companies are integrated into SAP right yeah, but That's just that just shows that that is I think a stark example of the difference Yeah, I mean I mean that clearly is slow. Yeah, I mean they should be instant Yeah, it's a long way to go for SAP, but an SAP is cutting edge. It's not like they're yeah, no, they're not Yeah, they're one of the first companies to even do that The contrast is just striking I slightly disagree though I think the the incumbent enterprise companies are gonna catch up it'll take some time and they'll they'll do it through Acquisitions and you're already starting to see it happen You know Dell is doing interesting things with open source right now, and you know, they just a few years ago I just like a year ago. They were thought of more as a hardware company They're still thought of more as a hardware company. They're doing really interesting things with open source now and It's just a killer be killed environment We were talking about the difference between traditional IT shops and cloud service providers I think I would agree with what Clint is saying the big guys is to buy their way in they got balanced Yeah, the old IT shops are gonna have to adapt to it that and in some ways Yeah, they'll they'll start getting cloud services or they'll start building private clouds and they'll use Puppet and VMware and and that sort of thing And they'll they'll have to catch up because they don't really have a choice and here's how we increasingly look at it It's just raised to the middle on the one side You have these you know the old enterprise legacy technology companies who are trying to be innovative They're trying to be young They want to show that they're they can play with the cool cats right the cool cats right they want to be Established they want to trust it right so they're both racing to the middle And so you starting to see companies from the old legacy world who are starting to innovate I would say HP vertical as an example that you know the ability to do data visualization with with a company like Tableau I would argue that sales force is is is really moving toward the middle too So I really think there's gonna be this race in the middle and who's not moving so that's a great point I mean because I think you're seeing innovation movement. You're seeing some tactics In that strategy, but who's not moving because don't move or dead basically is what we're saying So who are some of the players that you can say are not moving fast enough or maybe a little slower or actually standing still Can you name any or you know? I can't think offhand of anyone who is moving who isn't moving. I mean IBM You know, they're an old Giant, but they're you know, they're picking up companies like Natesa and Citrix is doing You know, they're pretty traditional, but at the same time they're they're acquiring companies. They're doing new stuff with virtualization VMware is just rolled out some new devop tools a couple weeks ago. We wrote about that on Silicon Angle I mean every company wants to be a part of that. I think it's a perception issue and I would I would point to Oracle You know to some extent and I and I think maybe Microsoft and to some extent too, but they're showing a lot of innovation as well What do you guys think of the story? So when we before we came up here We were talking about some storylines and that we were kind of watching One was the Hortonworks Cloudera thing and map are you know the different approaches any feedback on the hallways? And he buzzed about that as it kind of kind of watered down as it was over-hyped up or what's your observation from that? Big storyline, you know, you know, it's Cloudera's show and so people aren't talking about it that much I have been hearing about HPCC here and That you know as a complete alternative to Hadoop people are talking about it people are talking about Hortonworks map bar is here You know trying to give away free iPads or something. So I mean, this is Cloudera's show, but it's really not it's not Cloudera world Interesting debate should this be Cloudera's show. Yeah, I heard yeah I have one interesting feedback about Cloudera that it's some are confused about their marketing You know, are they a services company? Are they a consulting company? What about their distro? So where where are they focusing and how are they going to make their money? So we clarified that so we had everyone on at Albany's today We put that hard question everyone clear CDH is absolutely 100% open CDH is Cloudera's distribution for Hadoop now for the people who aren't students of this sector who were just new We've been following this Cloudera used to actually call their stuff Apache Hadoop what Apache did about a year ago They said whoa, whoa, you can't use the Apache name So what they had to do they had a big branding decision internally at Cloudera I think CDH kind of creates some confusion. It implies it what is what is Cloudera's distribution for Hadoop CDH But it's really still 100% Apache open source. It's a trademark brand issue with Apache So that's one thing that kind of got things screwed up Cloudera a little bit was you know They had to go out and private label private but branded them and they wanted to have some sort of branding So CDH is their distribution for Hadoop their distro as they say and that's their version It's a completely 100% open source. They're only charging for the enterprise suite which includes the management console and support And they charge for training and I'm sure they charge for the console but to be clear it said we are a software company So they're scaling the district the district. They're creating their own distribution That's a hundred percent open source that contributed to Apache and they contribute to Apache It's just their distribution of Apache Hadoop. That's their route to scale Their route to scaling is bundling that in and the NetApp deal I think is a telegraph of the channel strategy, which is Take our distro certified distro and I call that still open and then they're buying the enterprise suite All our wholesaling enterprise suite from Cloudera So NetApp is paying a wholesale for the Charge version and they get the distro. It's free so they can go with Apache Hadoop But they just choose cloud areas distribution that's it and now NetApp resells that pay the wholesale to Cloudera and then NetApp will wrap whatever services around it and That's essentially their go-to-market for the distribution As to your question about the narrative there it just I don't see it playing out much here People just aren't really people who are here just aren't talking a whole lot about Cloudera versus Hortonworks If it's playing out at all It's playing out in the question about whether there's going to be a forking of Hadoop and that that kind of plays over to MapR because their version of Hadoop is a little different and From from what Cloudera is doing with where they're just building on top. What is the map bar? I know you guys had a post on MapR just went up Clint I mean people are saying to me MapR has got a chance serious chance because EMC is backing it up It's just a proprietary distribution I mean MapR is a fork, right? I mean, it's not a fork. It's a really different product. That's what I'm saying But there's no Cloudera versus Hortonworks to me anyway is not a fork. Do you guys agree or do you see that as a fork? Yeah, Cloudera and Hortonworks are not forks. I think MapR does count as a fork I haven't talked to them about it. So I don't want to go out on them. John's right. I mean technically it's not a fork is in the community. MapR is completely rewrite Yeah, I just call it fork. I just call it something different. It's a completely different product It's proprietary MapReduce. Right. So I mean, I mean the amount of forking that's going on with Cloudera's management suite is you know I mean that's to me. That's not an issue. There's no forking because they're all contributing to a patchy See, that's the thing. Everything is going to be a patchy server. So now I just don't see the forking issue today I guess the fragmentation is more I guess what we What we should say rather than forking, but so let's handicap this thing because we've been doing it all week You've got you know the I've said the the guy who's number one is gonna make a ton of dough Number two is gonna make some dough and number three maybe make a little bit of money and there's no fourth place, right? So you got Cloudera you got EMC and in Green Plum and MapR and you got Hortonworks What do you guys see as the sort of position of each of those horses now if you had a handicap it I would also look at IBM and the Teaser too. I mean as a Hadoop distribution platform. Yeah, and I think they're a little bit different because they want to play from all sides They don't want to just use one that they're they're they're willing to work with multiple distros And I think that they've got a their approach I think is interesting because they want to make it dead simple that is their goal and IBM But then are they trying to so the of the three I mentioned the three I mentioned are trying to be the data operating system, right? Yep, the data OS the next you know the windows of data Right, so if IBM's doing that that's you know, let's put them in that camp But right now though, I think those are the three big I don't think I'm trying to do that I don't think so either right and IBM is the real loser EMC and real loser in this show because they're not here Well, we'll talk about that too, but Lexus nexus you would maybe put in that yeah Because they're not doing Hadoop, but they're doing something so similar that I I would put them in that race So smart monies on Cloudera. They've they've got a head start. They've got customers. They're resellers So, you know the smart money, you know, you're gonna you're gonna bet on Cloudera What works is so new to the market, but they have deep expertise But they're they're new so they're their player by default given their contribution and they've got a real shot at being number one, but You know Cloudera's got got the advantage right now just for by being first But we map bar has a has an uphill battle even with the emcee even with the emcee I mean my argument is that they have a product. They have customers. They got a distribution channel They don't have customers using map bar and green pump this very yeah, they're they're not out of the running by any means But I'll pill battles. What do you think compared? Yeah compared to Cloudera, and I think we're on agreement Cloudera is definitely the favorite Yeah, who's the second favorite Horton works? I'd say Horton, but here's a here. I would we were talking about this we were talking about this last night It's like hands on the odds. I guess Patriots last week and they lost so you know at 20 to 1 I'd pick emcee at 3 to 1 probably not Okay, go ahead, but I you know we're talking about this last night Who who do we see out there doing the smartest most elegant marketing? anyone well Messaging wise. I think Horton has been fantastic. So I would take that into consideration again I am sees marketing is fantastic cloud meets big data. I mean emcee John Furrier really started that but then you see Emcee marketing is the best right Horton works PR is the best for worse doesn't have any marketing They have nothing to market a cloud area has got a product red. Not yet. Cloudera is playing the Word of mouth we're in the community They're gonna win by having all the soldiers on their team I'm waiting for that super marketer to come out and really take a really and really and really Really take the market and say and you get that mind share and I just I'm not seeing it There's a lot of Cloudera Cloudera does not have a VP of marketing. They do not have a CMO They've been looking for someone for a bunch of months now. They do need one. They have PR now. They have a good PR team and Little bit confusing they've been they've been holding their message back, but now here they're messaged up I mean, I got to give Cloudera a hand on this one. They're messaged up beautifully. Yeah, they're hitting all their marks But their marketing is is simply we're just going to be good with our product and let the product speak for itself I personally think they need more than that. I think they need to play, you know The leadership role thought leadership is big. They're number one They have all the horses in the in the stable with amer bacher the data science team I'm a dollar Mike Olson. I mean, you know, Mike Olson is a Database guy. He knows that old business. He knows the new business. You got amer dollar. Who's a rock star Jeff? Hammerbacher we saw him on the Cube. He's fantastically smart. He understands data and their whole team is like that, right? So Cloudera has the horses that so I would bet on Cloudera Simply on the fact that they have the horses and they could continue to lead But I would tell you there's a screw up at this point Again, I would agree with that but I will tell you I would not count emcee out of this I was at a CIO emcee CIO conference a couple weeks ago one of the few Individuals that was invited that you know that wasn't a customer and I talked to a lot of customers and there are numerous Green plum proofs a comp proof of concepts going on out there now They didn't necessarily all include map bar, but I just I'd see them finally getting their act together It emcee salesmen knows how to sell hardware and they're packaging it as an appliance It's a good packaging strategy We saw that from from net app. So again, I would not count them out. I wouldn't either I mean, I think emcee this is one of the thing that I'm watching the thing that I'm watching in all this is The indifference of the customer base because there's a lot of diversity in the kinds of solutions that are going to be rolling out You might have people who love Cloudera in a certain area gonna have emcee customers that might have a different requirement So it's it's to me. I don't know what the indifference is to either switching Or not so yeah, to me that's the factor that I'm watching clearly there are advantages for everyone involved Again taking Cloudera side as a leader We can have a whole nother conversation about this, but but I really like IBM and you know, I'll just leave it Yeah, so so now and we don't and I definitely wouldn't count Lexus nexus I really like IBM Well now so why do you like I mean because I might be am I we haven't included them in this discussion link as I'm saying They're not going for the OS of the data, but maybe they are maybe they will come in Well, I think you need to look at everything that IBM's doing and their analytics approach I think it's really really smart. Hadoop is really at the base of their you know the base base this strategy Yeah, really really right there open source They've embraced other do and if you look at some of their their products that they have that they've developed such as like Info streams, right? I mean Very very good technology big sheets, you know, which is a wonderful sites, you know Which is wonderful technology for you know for really kind of reducing kind of that the sentiment of social analysis I saw one demo that they did where they actually the technology was used to pick Pushing boots as the number one film 18 months ago Based upon the sentiment analysis they did they correlated with advertising dollars They were spent by dreamers the amount of 18 million dollars Those are the kind of examples that I want to get that I think are really powerful But I think you also need to look at IBM's history of their of their of their acquisitions They have been focusing on analytics and there's a really good point Natesa SPSS, right? Cognizant Jesus and a rich open-source culture I mean I mean Steve Mills basically said hey I'm gonna spend a couple billion dollars to neutralize Microsoft's advantage and IBM did that I mean I agree IBM's got some good call. They actually are not doing good job You talk about marketing. They are really blowing it because IBM smarter planet Well, that's a high-level marketing plan that this company why but specifically the big data They've really got to up the up their messaging because they're really left out of the conversation I mean look at Alex, you know just brought him up. Okay. That's a good conversation. They're not even in there I mean they had their their event just two weeks ago and But it goes back to Mission on demand and then once that's over the conversation continues without them and But I think it goes back to the what I was saying earlier too. I don't in terms of marketing I don't see any clear winner. I don't see emc as a clear winner I don't see cloud era as a clear winner I think that it's so that this market is so young I mean, I think I think HP blew it at discover with Leo getting up there Talking about plumbing when he could have been pushing really big visionary things like big data I think that would have unsettled the market HP blue Well, they blew with cloud and they're also blowing in big data But they you know I talked to Colin Mahoney at Vertica last night at dinner. He was saying there's some big things coming out So I'm looking forward to seeing HP. But yeah, I mean, I think the the the jury's still out on who's going to win But here's the wild card for me Hey, you can use emc the example. I love that marketing campaign cloud meets big data. They nailed it very relevant The problem is is that in the open-source world that we're living in Community participation matters. So for example, IBM is not here emc is not here that matters So I think that's a big fat oracles here oracles here Netapp is here Cisco is here troopers not here. So, but you know, if you're not present you just do boy Wait, so you're saying de facto that Hadoop world. I mean, I'm asking this. You think Hadoop world is The most important big data event or one of the top three or five. I think and even though it is a user conference It's interesting I think it's I think this is the most important event only because the people building the industry are here And I've seen this before in other industry movements like client server The conferences that matter are the ones that have the participation of the people actually building the industry that is embryonic in big data Right. So I mean again last year this commerce size and makeup is different And the people who are getting excited are the ones like us who were president creation and are actually doing the work right so the in the sessions they're packed and They're asking thoughtful questions and these are the guys who really have questions and they're really doing some good work This isn't like a puffery, you know smoking the you know the peace pipe and you know payola conference Where and there's vendor and vendor spales going on Like other conferences. Yeah part of the question about who's who's gonna win I think you know we should maybe talk about who should win and I don't think that's clear yet who should win and I think that's part of Why we can't really answer who's going to win and who should win won't necessarily be who does win Well, but there it's really not clear yet who in the in the long run is going to have the best most mature product Is cloud era has the has the most mature Cloudera and HPCC have the most mature products right now, but it's still so early why I like Cloudera So that's a good point So who should win is kind of a good way to frame it so to me I'll just it's hard to say vendor at this point, but I'll say the approach that should win is an approach that provides ubiquity of a platform that enables developers to build products fast with Distribution and monetization meaning hey, I have an idea. I want to serve a market as a market for it There's demand I want to build a product and sell that product and make money. That's how it works in business, right? So I think the Cloudera messaging points to that their distribution deals With net app with SGI They're enabling the marketplace Ed Alvin is basically said we want to be out of the direct sales business. We want to go indirect I got to tell you there that that's a good story if they can pull that off That's worthy of winning and they're already in the lead for that. So, you know, so I still kind of philosophical But I still contend I wanted you guys could weigh in this at some point I don't know if that's three months six months two years Cloudera's got to demonstrate that it can transition all its goodwill Like Edmunds comm who's using Cloudera software to actual selling something Yeah, Edmunds is not paying Cloudera as an example and many many customers aren't so they have to figure out how to eventually get now I don't think it has to be tomorrow just raised 40 million they can hang in there for a while John I we're having this discussion It really wasn't a debate because we I think we agreed but at some point they've got to go from okay We're raising the you know the value of the company and increasing the awareness of the company to we got to start selling stuff What do you guys think about that my question is why do they need why did they need 40 million dollars? Well, I think I My answer would be because they're up against guys like EMC and IBM And they haven't started scaling The sales of their product. Yeah, they can sell. Yeah, they're investing in sales and marketing and reseller partnerships mostly I think I think it's an upround and so okay So I just saw that Mike Olson's in the house He's gonna come on the cube for kind of a wrap-up, but before we bring on Mike I want to ask you guys just been doing a great job Covering the event content's been fantastic Post event what's on your radar? What is what's the action items for you guys after this event? What's any kind of epiphanies here and I'll see we talked about dev ops. What else is on your radar? I'm gonna take some pictures of the chickens over the weekend, you know Have some eggs Gamble the eggs. I think this we're coming into the end of the year. It's gonna be fun to kind of look at this past year Looking at some of the you know who are the top players and data analytics Who are the top players who are really kind of you know building those developer communities? Who are the ones who really make the play and in 2012? I that's I think gonna be the focus for the rest for the next several weeks Yeah To be I guess a little more specific big top is something I'm that's on my list to look more deeply into I Need to do spend some more time looking into are in revolution analytics I think that they're you know, they're doing really interesting things They're they're on the edge of of what data scientists are actually using in order to do work with big data So I so I think that those are two particular areas Data visualization if I'm gonna be particular I'd say you know, I'm still working on this post on you know Is this to do ready for the fortune 1,000 and this this event really kind of helped me Yeah, I think we got to live by some ideas. I think the mission is clear cloud meets big data You know for us, it's cloud virtualization VMware that's gonna play into it I'll see the Hadoop we got cloud air and all these guys and analytics so all right guys Thanks for coming on appreciate it great event We're gonna wrap this up, but but appreciate it good work job this week great job Thank you. Okay. We're live in New York City as there are top guys the bloggers out there doing the Scouring getting stories talking to everybody getting briefed to Alex Williams and Clint Finley silicon angle comm exclusive