 Okay, we're back live here in San Francisco, California. This is Oracle Open World 2012. Exclusive coverage on SiliconAngle.com. The Cube, this is our flagship program. We go out to the events, they extract the signal from the noise and it's exciting here. It's shut down the whole streets of San Francisco. 50,000 people. I'm John Furrier, the founder of SiliconAngle.com and I'm joined by my co-host. I'm Dave Vellante at Wikibon.org and we're here with Jay Kidd, who's a senior vice president at NetApp, a company that he helped rename from Network Appliance, which our previous guest, Vu-Win, referred to NetApp as Network Appliance. Old habits die hard, but the new name's sticking and it flows off the tongue, so welcome to theCUBE, Jay. Thanks, Dave, we're really going to be here. Good to see you and it was good. We were trying to get Vu-Win to talk about NetApp a little bit more detailed, but of course they can't do that at NASA JPL. So you can tell us how great NetApp is. So we're here at Oracle Open World. It's kind of an interesting event. We were at VMworld last month and different vibe, very open, everybody's friendly. It's all about the ecosystem here. It's about the ecosystem, but it's a lot more competitive. But interestingly, NetApp doesn't seem to be in the line of fire directly anyway of Oracle, which I guess is a good thing. Tom Georgian says that Oracle just likes us less than others, like this is how we put it. But still, I don't know if you can say it, but I can say it, Oracle's definitely a customer of NetApp and has been for a long, long time. And even though it's probably migrating to more sun storage, but so here we are. And what do you think of Open World and what's NetApp doing here? You know, your comparison of this in VMworld is really good because VMworld has become probably the biggest infrastructure show. And I see Oracle is really becoming the middleware, the application, the database. Oracle is such an enormous company with such a tremendous impact on the industry. It covers such a wide range of things, but it's nice to be out of the line of fire for Oracle every now and then. But we even have been doing some things with their 12C announcement. We did some work with them for the integration of 12C with our cloning technology to let you actually clone the component databases, those pluggable databases that plug into the master. So we still got a great relationship, a great partnership with Oracle. And it fits maybe into the multi-tenant aspect of 12C as well, right? Or just simplicity of upgrade of development if you want to clone a test dev copy of one of the component databases. The key thing is we still work really well with Oracle and they like working with us. Jay, last year at Oracle Open World, we had a chance to get some candid footage of Tom Georgins with Moneyball. You know, the author writing, signing the books. Obviously big data last year was a lot of hype. Now it's becoming more mainstream. Can you just talk about the perspective of what's happened with the NetApp and your view of the industry just one year later around big data. Obviously the database thing you're mentioning is a big part of that, the in-memory and the advancement. So what's changed with big data over the past year? I know you guys have been looking at it before that. You've seen some changes in your customer environments in terms of preferred storage solutions. That's non-sequel and so on. So what's the change in one year? Yeah, so a couple of things are going on with big data. Well, first it's now ascended to the most misunderstood term by our customers, replacing cloud. Partly because cloud's got a little better understood and partly because big data's on the rise. More people are engaging in it trying to figure out what it means. We work a lot with big data of different types. The thing we found most people really need is they need infrastructure which can respond easily and with a high degree of agility to the inflow of data that they may be getting. It could be transactional data or it could be interactional data but they need to be able to capture it and figure out how to make use of it. So we were talking with Billy Bosworth, the CEO of Datastax and he says, the DBA is not going to go away, it's going to change to data management expertise. Not so much administrating, more of a management, more of a broader portfolio. Analytics and automation are two of the hottest trends that we're tracking in the data infrastructure space. You guys talk about it as agile data infrastructure. Can you talk about that? One, the roles within the infrastructure in terms of your customer base and then two, analytics and automation and those two trends in particular. You're seeing a lot of, I've even ads in the papers and online now, the papers, dating myself. But for data architects and sort of data engineers, people who can figure out how to look at a stream of data and figure out how to structure so you can derive information from it. That's a theme that I think it's creating some new types of roles. It's more than just administering a database, it's understanding how to derive value from chaos. The infrastructure side of it though is all about how do you build scale? How do you build infrastructure which is always on? Because these big analytics applications, they need infrastructure which can come and go as it's needed for the different jobs, but as a whole, as a pool of infrastructure, you can't ever really take it down. Analytics are becoming too important part of the interaction with the customer to do that. You guys had a marketing positioning in a company view a couple years ago, data efficiency or efficiency storage, efficiency's always been kind of a big buzzword as the environments are kind of consolidating and then expanding again under these new architectures. Efficiency still is a big part of the equation, right? So you can talk about some of the new challenges around dealing with efficiencies when one, big data's the most misunderstood term on the planet, and two, there's real cost advantages and performance improvements around, say, flash and these new architectures. Yeah, so efficiency and flash go hand in hand with each other. Efficiency just in the amount of data that's streaming into companies or organizations or governments anywhere today, they have to figure out how to capture it and how to save it as cost effectively as possible. Everybody knows budgets aren't going up at all, much less at the pace of data acquisition. So efficiency is actually critical. A lot of the analytics tools out there require you to store multiple copies, three, four, five copies of data in order to protect it. There's much more efficient ways to do that. So that creates an opportunity for delivering analytics at a lower cost point. Flash is changing everything in enterprise storage. It's still too expensive now to be all of the storage for an application, but it's reached the point now where a little flash integrated with a lot of, say, the disk creates a really powerful combination. So Dave and I always talk about NetApp and living in Silicon Valley. It's fun to watch NetApp as a startup emerging to be this massive powerhouse. And now you guys are the, I call the large, independent kind of storage vendor. Everyone else, like EMC, they have all kinds of broad portfolios. You guys are so resilient. You always seem to reinvent and move at the pace of the business. Can you comment on one, NetApp as a company right now in terms of the health and internal feeling in the culture? And then two, the direction you guys are going to take with respect to some of these new changing plans because storage, it's an interesting spot that you're in with Flash and with the growth of cloud and mobile. Yeah, one of the things that sets NetApp apart, and I've been there now about seven years. My history with NetApp goes back even before that. The culture of the company is a major competitive weapon. From the outset of the first orientation to the behavior of the executives to our history, we tell our people that their job is to take initiative, figure out what needs to be done, figure out how to make their job better and how to make the success of the people they work with happen faster. But above all things, we're committed to the success of the customers. So we listen to the customers, we understand what they need, hard to figure out how we can deliver for our customers what they need. That will allow us, it allows us to weather storms, to weather technology transitions. We pioneered unified storage 10 or 11 years ago now and that became very popular. We really drove sand into a company that was religiously on the NAS bandwagon. And now sand versus NAS is sort of yesterday's argument. It's all just part of network storage. Storage efficiency, integration with VMware and Microsoft and the whole virtualization trend. And now our scale out architecture with Clustered On Tap, we've been able to continue the pace of innovation and that really comes as much from the culture of the company as it does from anything else. So NetApp celebrated its 20th anniversary this summer and now you haven't been there 20 years since you've been there seven, but you know well the history of the big bets that you've made. And I wonder if you could maybe double click on the last point you made about clustering. So you guys bet big on the internet, the dot com that worked and then it blew up. Then you bet on the enterprise. You had to make that transition and your founder, Dave Hitz, co-founder, loves to talk about that. And then you bet big on VMware. That paid off. You were able to, you know, buy our data and Wikibon, we've studied this intensely, keep pace with EMC who owns VMware. You guys would argue that you actually integrate better with VMware than anybody else. Yeah, well you've got a strong argument. And now you're making a big bet on clustering. Talk about that and scale out. Talk about that, why that's so important and why you'll succeed. Well it comes back to what we were talking about before with the pace of data acquisition and data creation in companies now is so high that they can't afford to throw it away. So they have to figure out how to manage it. On top of that, they're consolidating infrastructure at a rapid pace. So VMware allowed people to consolidate servers at a 10 or 15 to one rate which consolidated data centers in turn which drove to a higher consolidation of storage. The net effect is most of our customers now have a smaller number of much larger points of IT and therefore those larger points of IT have to serve a much larger base of applications. They have to always be on, they have to be able to scale inexpensively to start small and grow to infinity and they also have to bring a lot of intelligence to deliver the efficiency and the range of workloads and applications the net has always been known for. So one of the things that we always talk about on theCUBE and at SiliconANGLE is this role of the operating system. We talked about the internet operating system, storage operating systems, data center operating systems and we're seeing a trend where storage has become such a pivotal role in these converged infrastructures or data infrastructures or software infrastructures, software defined networking or whatever that the role of data where storage was involved is becoming the epicenter of the heart of this new OS not just storage OS but the data about data is really critical and that's going to affect network virtualization and other kind of cool new emerging areas. So the question I have for you is this is going to impact a lot of traditional markets that you guys played in, obviously in verticals like healthcare, finance, other big verticals and then also the data warehousing and business intelligent markets which are we're slow moving markets the pressure of real time and this convergence with storage at the center of the operating system you look at VMworlds they're all like operating systems pool, abstract compute. So I want to get your perspective and netapp's perspective on this trend data warehouses is business intelligence it has a long tail we talked to some folks earlier it's always going to be around it's not going to be going away soon but yet it's under massive disruption how does netapp and the storage business get in there and move data warehouses intelligence to the new environment faster a modern infrastructure if you will. Yeah, there's a couple of these trends that are going on in data warehousing and analytics the world seems to be dividing a bit into the long running deeply complex looking over big trends of time analytics to do more sort of background analytics and the real time gutter decide how much can I get what kind of decision can I make in 40 milliseconds to be able to respond to the user and those two types of applications are different now over time we think they will start to converge but there'll be more detailed analysis you want to do that takes longer and there'll be things you want to do faster with more data so Moore's law will allow more technology to be applied to both of those application types the key role that storage has to play is it has to make the data available inexpensively sort of a capacity tier for those long running analytics and then available instantly in an IOPS tier for those rapidly running analytics that analytics would be like throw it over there work on it and then real time is it's mobile or and you want so you'll never be able to afford to have all your data in the real time storage infrastructure because you'll always want to have more data than you can afford to put there but flash is creating infrastructure to allow much more rapid response so architectures which integrate flash with hard drives allow you to manage the data the most effectively to put the cooler data where you don't need it the hot data where you do bring that back to NetApp for the folks that might not be inside the ropes of the NetApp how does that relate to the current offering so NetApp is we've been working with flash integrated with our FAS our fabric attached storage controllers for about three or four years now we have SSDs you can incorporate in for to accelerate the performance of drives we have flash and PCI cards in the controller to speed up read performance we also have software that runs on the host to use flash that exists on the host to accelerate reads even faster for the applications that are running in the application hosts so now you've also made some partnership announcements you've made some big announcements of your own so you're building that flash stack out pretty substantially can you talk about where you see that going so in the long run we see flash existing as close as possible the applications but because it will always be more expensive per gigabyte or terabyte than hard disk you're going to want to optimize the use of it extensively so we believe the right model is you want flash close to the application on the memory bus of the servers or close to the network but it's cash and it is dynamically loaded and managed from the backing store of hard drives which have all the data in a way that gives the performance you need at a much more economic point the last thing somebody wants to do is spend money on flash to hold data you're not accessing cold flash is a sin cold flash like cold dead fish I've made the point a number of times with somebody's marketing I've made the point a number of times that the function has moved out of the host into the sand and now sort of moving back ONTAP has been the point of management but that a lot of that function that data management epicenter and made a lot of sense how does that flesh out with flash can slower storage manage faster server close to the server flash or does something have to change there and so what does that mean for you guys yeah the key thing is with flash close to the applications it's easy for a backing store like a capacity tier like an ONTAP system to easily load a large amount of data at a time faster than the applications that are consuming it are going to read because the applications are going to read randomly so you just want enough of a working set in memory in the application host feeding that from the storage system just a way to make that work well when you had a tiny bit of flash in the host it was too hard for the storage to fill it quick enough in response to the application you could put a terabyte of flash in a host that's a pretty big working set but most applications will work fine with a terabyte or four terabytes or six terabytes of flash so then it just becomes how do you protect it the last thing anybody wants to do is have to back up a thousand individual servers or five thousand individual VMs in each having their own discrete process you want everything up at the server side to be stateless so you can turn it on or off if it breaks down you can run the apps on a different server and all the data managed in a central way that's really what ONTAP is driving Jay, talk about Larry Ellison we love Larry he's been, he's a statesman, industry legend he's a tech athlete as we say and he's one of the old guys still punching Joe Tucci out there on the east coast and Scott McNeely is pretty much retired he's doing some side deals and son's playing golf I read about that on Facebook all the time but Larry's up there he's punching, he's a showman what do you think of his Exadata presentation I mean that was pretty aggressive I mean he slams EMC so we'll see how Joe Tucci responds so you see that out there what's your comments on his mojo and maybe he's oversimplifying it but what's your opinion on his presentation so Larry's reputation in the industry and he has done a phenomenal number of things in the industry Oracle's a company of tremendous power and influence there's been no secret that Larry wanted to own his own data center agenda, infrastructure be able to write the script for a company for their entire data center infrastructure the whole sun acquisition and the Exadata play gives him a box with which to do that where we disagree and most of the customers I talk to is I ask them what they think of Oracle's full end-to-end system agenda and they say well if my applications are Oracle I'll look at Oracle hardware if they're not I won't so I think that it'd be difficult for Oracle to own the whole data center with the agenda they have but then you see the other things that Larry doesn't talk about quite as much of Oracle partnering with people and doing the right thing for customers across a much wider range of workloads than Exit serves but it's entertainment and it's interesting I mean we've commented earlier this morning that he wants to be the Steve Jobs of the enterprise, Apple of the enterprise and you know Apple doesn't own my home I have cable TV so they don't own my home so I don't think enterprise is who I mean I don't think anyone's ever been successful as the one-stop shop going back to maybe IBM back in the 40s, 50s so it's a multi-vendor world exactly so maybe in a silo Larry's got a baked out solution but the reality is CIOs have to deal with legacy and multi-vendor so given that how does a CIO and their team of developers look to the future given the noise of what's going on around them well I think there's certain trends which are secular across old and new, big and small the virtualization of servers is widespread the consolidation of storage and we believe ultimately the scale-out of storage will become widespread as well the evolution of networks to be software-defined networks I think it's a widespread trend as well what CIOs need to look for is trends which are adopted and embraced by more than a single vendor in the industry as you know a single vendor is a rogue story three or four vendors leading a trend is going to create enough choices for a CIO to make but still push the technology forward this industry has not succeeded on the innovations of a single vendor on the way we believe and we've been commenting that we believe that there'll be a data center operating system a lot of multi-vendors, a lot of subsystems a lot of coordination, automation I think your data infrastructure is a cool positioning so that being said the future seems to be in the software-defined networking because that's like the last bottleneck of convergence it's like the last stop we've got servers are done storage is looking really good off the tee right now now you go back to networking Cisco, Juniper, and now VMware has Nacera So, yeah, NetApp I mean you got to look at that and say hey we know software we know a little bit about hardware and moving data around what's your take on one software-defined networking software-defined software virtualization or network virtualization and how does that affect you guys? So we're a participant in the data center but we're really not caught up in the evolution of architectures in networking we work in layer two networks layer three networks and we can be found in the network and where the data resides I've had a few people asking about software-defined storage and I point out that, you know That's you Yeah, we already do that Well, we already do that but the one thing in the data center that actually still does work when you turn off all the power is the storage it holds the data when you turn it back on it's still there so that has mass it has gravity it's the Higgs boson of the data center architecture, right? So it doesn't float around quite as readily as you'll see in some of the other architectures so NetApp is focused on storage and data management we play well with the networking vendors we play well with the application vendors great relationship with Cisco with VMware, with Microsoft we're happy with our position in the market I think we can grow from there So, but since I've been in this business I've been told that hardware's going to be commoditized and here And we've been in the business a long time I've been in this a long time you can tell my mic right here having said that with the consumerization of IT all this software led networking software defined storage you know, Hadoop you're getting to the point where there's a major push from companies like Oracle and VMware maybe even to commoditize hardware even further at the same time you see gross margins, you know, aren't they're pretty steady you're maybe not inside of Oracle for the hardware side but they'll creep back up what's your take on that, Jay? I mean, custom A6, not custom A6 you see it in both places is it more the same is there legs to that commoditization of hardware and what does that mean to NetApp? You know, you hear about commoditization of hardware a lot you almost never hear about commoditization of software Well, open source commoditizes software, right? Not really, there's a lot of innovation and open source and it may lower the cost of deployment that's not equal to commoditization Commoditization is the sort of the absence of innovation and I think software's got a long way to go NetApp is first and foremost a software company that we apply our software to the most cost effective hardware we can get and we do amazing things with efficiency of storing data so I think commoditization is a trend for disk drives and for some semiconductor parts but I wouldn't really say it's a trend in the storage industry or the network industry or the application industry And the enterprise in general I think lower cost will always be a trend That's table stakes though, isn't it? That's table stakes, absolutely agree with you Commoditization implies a lack of innovation I don't think that's the table that people want to play at You were talking about earlier about IT budgets and the discourse in the industry for the last 10 years really since Nick Carr wrote this IT matter has been do more with less IT cannot bring sustainable competitive advantage with the big data discussion there seems to be a change in that sentiment you're seeing real examples in use cases of analytics and big data applications actually having a fundamental change in the productivity of organizations Do you see in our lifetimes when CEOs start to say okay we have to invest more in IT as a percentage of revenues have we cut to the muscle and will we bounce off the bottom in your view? You know I think CEOs will look at that and say I get a tremendous amount of value from a few innovative things done in IT and how much did they cost? That's all? That's not where most of the cost is 80% of the cost in IT is in keeping the lights on and legacy applications and I don't think that these innovations that are a great competitive advantage are going to be an ever be legacy they'll always be innovating which is it comes back to sort of where we started of the whole agility of infrastructure and the agility of data people need infrastructure that can respond quickly that can change quickly and that they need architectures that can change quickly to allow them to take advantage of and create competitive advantage based on the data that's flowing in that's never going to cost that much to do to create that innovation if you're intelligent and efficient around what you do with the infrastructure you have So follow up on that is to your point the biggest cost you said is the legacy infrastructure it's the labor to keep those lights on I mean it's probably two thirds of the cost of what we spend spending not revenue but spending is in labor do we have to solve that problem before that vision just laid out can occur and what are we doing to solve that problem? You know I think you look at the labor cost in IT and it has gone down as a percentage actually it's going up as a percentage of it because the cost of the hardware has come down but the amount of productivity and how much work each individual does it's still gone way up I think if you have a static static environment where nothing is changing you can worry about lowering the cost of labor so if a company isn't growing at all do you worry about lowering the cost of labor? As long as things are changing you're taking advantage of new infrastructure what's happening is the labor is shifting it's shifting from people who just run hardware and keep hardware running to those innovators who work on new applications that run on the big data analytics and on Hadoop jobs and things you shift the labor to there so it's a skill shift in labor rather than an absolute reduction labor What does that mean in your view to a storage admin and a DBA about 2,000 how does that change their role? You used to look at the very distinct roles that the server admin, the DBA, the storage admin the network admin and now we see a lot of organizations that have shifted to managing a vertical stack you want to provision an application with a golden image of an Oracle database loaded in a VM deployed with storage with backup protection that's all done by one team and the expertise across the domains spans across everyone on the team that kind of rapid provisioning of an entire infrastructure stack that's something that we've worked on with FlexPod and our technology at NetApp but bringing people together to do that as a team is what's going to change the way that the labor force in IT looks in the future So Jay, how are you spending your time these days when we first met you were running marketing at NetApp you went through the major brand transformation which has been very successful the Storage 5000 was sort of under your watch there another really successful initiative that you guys propelled and then you went into the product roles what do you do in these days? So my passion has always been the technology and how NetApp can bring innovation that matters to our customers so I've been spending more and more time on that looking at what we do with our management products our clustered on tap and the scale out architectures and I'm just looking, I spend more of my time on what we're doing on what's our overall innovation agenda and I'll be digging in even deeper to that in the future on the product side or field side or more on the product side what we build, the products we build first and foremost we're a product company with a great go-to-market team but the innovation in our products is really the soul of the long-term growth in NetApp on that, we love products we love product conversations so I have to ask the big data question because we had the data stacks guy and we've always talked to Cloudair, Hortonworks so the Apache, you obviously open source is great scale out open source, rah rah, it's doing great however the reality of servicing scale is a challenge operationally in all the efficiency things so what's your plans for Hadoop or NoSQL or is it Mongo? We have Mongo guys coming on too later today obviously there's interest in having a schema lighter a somewhat non-complicated schema store for batching or real time and that has to, it's going to be part of the portfolio at some point, not replacing anything yeah you may lose some business here in there but for the most part it's going to be a new element how do you look at that product-wise and what's NetApp doing? So we often look at, and I think what you're asking about is do we need to get into the database business in some way and have that as a piece of our offering like EMC did with Green Plum and our belief is we don't much I don't think so we can serve interest there's a lot of database players out there I just partnered with a few of them I think you'd be good on that front that's what we're doing we're partnering with Oracle, with Microsoft, with Mongo with Hadoop, with Hortonworks we believe we have just outstanding technology for storing data and managing the data that it's stored that doesn't mean we have to get into the database business or move us back open source is great it incubates the market and Lena Oracle is great at doing what big companies do is they see it across the chasm they put their blanket over it we now have that too so you guys bring a lot of industrial strength products to the market management mentioned other things so with that, what about the big data I call it the mushroom patch a lot of mushrooms are growing and some of them will be ripe, some will die so you've got to look at that so what are you looking for in particular for breakout products and tech from that area? so in the big data side we developed solutions to work in the Hadoop world we're looking at what we can do inside Hadoop I'm not going to say too much about it here in math what we can do in that open source world to allow Hadoop infrastructures to run more efficiently and with a closer and more cooperative linkage to the backing store that may be in the archive of data that may not be residing on the Hadoop grid at that moment there's a lot that can be done with scale-out architecture in that area so we can look around we talked to Val many times in the queue going back to some math studies a couple years ago so we know it's on your radar I just want to get your perspective and share that so I'm reminded John of the interview we did with Jeff Hammabocker one of the founders of Cloudera he was talking about when he was at Facebook one of his missions was to break the shackles of the container you know now in a way you guys you're into the container business right now having broken the shackles of the container business you know let's call it Hadoop the problem is it's not robust enough for the enterprise so does that you know exploded container sort of change shape and become more robust or do we go to a more containerized traditional storage world in the Hadoop if that makes sense it's a good question and one of the tensions in storage has always been between agility and flexibility of the data you need to work on versus protection of that data and protecting things tends to still be of there's a container mentality to protecting things either a volume or a tape or something I can go and recover from the architecture that protection is done by just making multiple copies through the infrastructure it's sort of the equivalent of backing up your personal data by emailing all your files to your friends it works but when you get something back it's pretty unpredictable how soon you're going to get it back and the same thing exists in the Hadoop infrastructure so enterprises they want great analytics but they want predictable performance particularly for these real-time type of analytics so you got to figure out how to make the infrastructure predictable yet still protect the data within it and that may be two separate problems that needs two different architectures Jay Kidd with NetApp Senior Vice President Agile infrastructure, data infrastructure Agile programming, if you don't understand what that is Agile infrastructure is kind of the same concept we didn't have a chance to talk about applications and data is kind of where apps kind of come together whole another conversation opportunity thanks for coming on theCUBE, appreciate it we'll be right back with our next guest after this short break