 Hi Vaughn, Dave Vellante. Dave, nice to see you. Good to see you. Come on in. Tight squeeze. Yeah, it's very tight here, but we thank a few logic nonetheless. So Vaughn, Oracle Open World, you guys have had a long history in Oracle. Changing. Absolutely. History, right? They're a customer and have been for a number of years. Of course, now they're a competitor. So you guys are like frenemies. Oracle's a war with the world. We don't know that. I wouldn't say competitors. The reality of the world is all your partnerships have some form of overlap. We're all working, I think. It's fair to say that we're all working in the best interest of the customer. Did you not compete with Sun? We're going to compete. I mean, we're going to have areas of overlap. A quasi-competitor. That's the way it is this day and age. Co-op petition. That's right. As I say, John's word is frenemies. Yeah. Anyway, let's start with what you do as a virtualization evangelist. What does that mean? My role inside of NetApp is what I like to call a bipolar role. I spend a portion of my time evaluating technologies, solutions, market initiatives, and bring my assessment of these items back in-house to an engineering group and ask them to advance our solutions, points of integration, write our value in this space. That's the more technical side of the role. The 180 degrees from that is the public evangelism piece, which is engage the public, customers, partners, help them to understand what are we doing in this space, making them aware, raising the awareness. And frankly, I think my greatest asset in both of those roles is I have a perspective that's not storage driven. It's much more focused around business goals down in the storage and how the storage capabilities can enhance the ability for one to achieve those goals versus a storage perspective that's coming back up into the stack. Yeah. And I think NetApp's actually, I give NetApp a lot of credit. It's always done a good job of talking the business value and not necessarily talking speeds and feeds. I mean, NetApp can talk speeds and feeds with the best of them. But by the way, I saw in your booth today, you guys got Billy Bean at, is it two o'clock? Two to four? Is he going to be there the whole time? So Billy Bean, Mr. Moneyball, Oakland A's, very famous book, Moneyball. Can we get him on the Cube? Can we get him on the Cube, John? If anybody can do what you can. Make it happen. So check out the NetApp booth. We'll try to get him on the Cube. That would be fantastic. So again, another very interesting spin on business value. I presume he's talking about analysis analytics. Absolutely. He's actually much more technical than I think most people give him credit for. Oh, I've inferred he's pretty technical. I have no idea. Kind of a math geek, I thought. But anyway, or at least appreciates math, I should say. So I want to talk about those two roles. The one where you essentially have to, first of all, prioritize and then sell internally on those integration points. Where are you getting your best feedback? Is it customers? Is it partners? Is it the development community? Where's it coming from? Yes, yes, yes. But talk about those. Sure. So first off, great question. Whenever you think you haven't heard a new and creative question, thank you for laying one out there. In the short answer, it is yes, yes, and yes. I think NetApp affords me this opportunity to have these various channels of outreach to receive this feedback and be able to mull it over, boil it down to its core, and then map that back to our technology to see if there's something that we can create. Sometimes some of these initiatives are obvious when we initially deployed data deduplication in the virtualization space. The mindset internally was this is for backup and archive. This isn't for production use case. From there, we drove it into that space, not just successfully, but it's led to enhancements in the IO, that the array processes by having deduping the cache, caching algorithms, and we continue to expand that type of advancements forward. Today, right, that's a long time success and one that we're noted for. But today, looking at startups, seeing who's doing what in that space, frankly, I still think a lot of the storage vendors are looking in the wrong space to have the greatest value in this shared virtual infrastructure, but that's my opinion. I think storage constructs really need to evolve a bit beyond the concept that we have today, which is SanNAS or Tomorrow Object Oriented. I think there's a lot more value that can be delivered in that space. Value comes in the format of points of integration, the stickiness and the ability of the platform to enhance the capabilities of the application itself where the infrastructure is one component. But where these sources come from? Technology partners, channel partners, customers saying, you know what I would like to have? They're all fair game and they're open sources to have creative thought. So you mentioned NetAppDdupe, which was an early innovation. I think you used to call it ASIS, if I'm not mistaken. What was the contribution, if any, that your team made to that innovation? Or was that just an example you chose that didn't have anything to do necessarily with your efforts? I'm curious, I want an example. Okay, perfect example. I'll actually even expand the scope. I think it's fair to say around 2006, with the release of VMware's VI3, notions of production use of data deduplication, or even NFS, was not necessary that everyone was on board with. I think there may have been folks who thought with the admin and server virtualization, it means that we will likely see customers deploy on fiber channel, that that's the enterprise format of a storage network. So our team led the charge in both of those initiatives as well as other areas such as plugins and integrations inside of vCenter, by actually vetting out the value prop, presenting it internally, we would typically get challenged back to show points of validation. So performance numbers, market drivers, try to align advancements in the market that maybe weren't crystal clear because you didn't have results, but the concept of dedupe really mapped well to the concept of driving up your hardware utilization at the server layer, right? So we could store more data per physical terabyte than what was normally possible through dedupe. So is that storage virtualization? Sure it is, but does it map to your goal of running more applications per server blade than what you could with physical? Absolutely, so we had to draw some of these analogies and win over, believe it or not, really, really smart engineers who make the code happen, but make them, and technical directors and things of this nature, allow them to understand where we can apply the use of some of these capabilities. So you have to do internal and external? Oh, you've got to sell, yeah, you've got to sell. Let's talk about the virtualization space. We were at VMworld a few weeks ago. We had Tom Georgins on, a number of other NetApp guests. You guys are doing a great job in VMwareland, no question. We at Wikibon have done a number of research studies on integration. You guys are right there with EMC as the leaders, as the perceived leaders within the customer base and the actual execution leaders. But there are others. Absolutely, absolutely. I don't want a three par. I think it's doing a really good job. The Dell Equalogic guys at Hachi is doing much, much better than they have. But EMC and NetApp stand out. Why is it, how is it that you guys were able to be so successful? You saw the trend coming. Were you just naturally following the market? Talk about how that came about, and then we'll talk more about where it's headed. Well, first and foremost, I really value the research and the reporting that Wikibon did around the integration points. I thought that was a very large amount of data for you guys to review, assess, and then come forward with your opinions. And so I appreciate it. I think it was fairly accurate. Fairly accurate. I think it was accurate. I shouldn't say fairly. I'd like to see you guys continue in that space, maybe on some annual basis or something like that. Look back and continue to follow that. In terms of where NetApp is in the integrations and maybe even not speaking officially for EMC, but I think NetApp and EMC had some individuals that saw the change in the infrastructure in the form of connectivity and viewed storage as no longer being confined to the array or the hardware platform itself. It's really an end-to-end conduit or construct, I should say, and the conduit includes the network, includes the virtual disks and the hypervisor. You've got to run in an optimal environment. The ability to coordinate activities through the entire stack down to and map those to the capabilities inside of your platform were really opened up and exposed by VMware. For us to move forward and then make tools that allow us to provision or have application-aware backups or things of that nature really became a no-brainer. Once you understood that the paradigm had shifted from only being an array and that the demarcation point of the storage object was actually this file that the hypervisor was addressing. Again, I just think it's looking at storage a little bit differently than if you're just looking at the protocols and the chipsets that operate inside of your platform. I've got to ask you this question. We had Tom Georgins on. We had it on a number of times. He's a great guest. He always has really good insights. He's very smart. I always have to ask, oh, EMC owns VMware. How do you compete? Blah, blah, blah. I have to ask it. Absolutely. It's a fair question. He always has a good answer. The one that he gave last year at VMworld, he said it probably better than I'm going to, he said essentially that EMC is not going to sub-optimize its investment in VMware to try to sell more storage. Absolutely. I thought that's true. I've always said VMware is going to be more valuable than EMC someday. Smart people, credit to EMC for buying VMware. A lot of people wish they had. Then we started thinking about some other things and I don't know if this is Machiavellian on our part or not, but there's a scenario out there that was put forth by some folks in our community about what they call the storage cartel. So essentially you have EMC and VMware and obviously there's insider action going on there. But then you've got EMC essentially saying, yeah, go ahead, VMware, you need to be open with all the other competitors because we'll sub-optimize your value. Well, you don't do that. So there's NetApp and there's HP and IBM and all those companies sell a lot of VMware licenses. But the scenario is that there is this sort of cozy relationship forming amongst the four or five large companies and of course the idea is prices stay high. That's what a cartel does. And it's potentially risk stifling innovation long-term. Small companies can't get the SDKs for things like VAI fast enough. I'm sure you've heard this scenario. But you have to ask that question because as a software company you would think VMware would want to commoditize the hardware and make it as inexpensive as possible. I want to give you an opportunity, there's a NetApp representative to respond to that. Is that a wacky scenario? Is there any merit to it? What are your thoughts on that? So great insight and interesting thoughts that are being brought up here are suggested. From my perspective, frankly, I tip my hat to EMC. They are in a very difficult position trying to ensure that their investment can mature and they need to stay out of the way of the channels that allow VMware to be sold and delivered to their customers. At the same level, I have a lot of respect for how VMware manages this relationship with one of their principal investors. They, for lack of a better term, have a Chinese firewall in place and allow the two entities to be operated as such. From our perspective, I'm actually very happy with how aggressively VMware seeks to pull more storage, what has historically been storage-centric attributes into the hypervisor level, right? So whether we're looking at IO control, for example. Into its stack. Yeah, into its stack, because that allows us to then work more on coordinating at their level, at a more intimate level with the application itself. So it allows us to advance our solutions and then we can focus on competing in the storage space on the merit of our capabilities to plug into the frameworks that they're creating or on the merit of that unique capabilities that we have inside of the platforms. So that levels the playing field. Absolutely. So I think... For those guys that get the SDK day one. Yeah, so if I'm a startup or a smaller storage vendor, if I figure that I'm going to plug into, say, a vStorage API of one of the many different programs or the array integration of the data protection, et cetera, I probably should consider that that's a pretty level playing field, that those APIs are going to have some form of universal appeal across all the storage platforms. Where you're really going to exceed as a storage company is what can you do in terms of innovating and delivering a product that's outside of those APIs that has the ability to be integrated into the virtualization space and provide value. So I'll give you an example, and I know this would be a NetApp-centric plug, but we've had a construct of a virtual storage control or a virtual storage controller. It's a means to partition physical hardware and have multiple logical or virtual arrays, and each array can consider its own security realm or domain. But we've been able to leverage that construct to release enhanced secure multi-tenant designs or to enhance the non-destructive ability to move a workload between controllers, and we're going to continue to go down that path because what customers are telling us is that they see a tremendous amount of value of encapsulating the storage access to a logical entity and no longer binding it to a physical entity, and now they get to manage their infrastructure assets, an array platform, much like they would manage a virtual machine sitting on a hypervisor. I love these scenarios because I'm an analyst and we love to put forth these controversial statements and things, but I think that ultimately that, what you just said is the key is to deliver business value. And ultimately VMware has to keep continuing to deliver that business value because it's got competitors as well. So that's why I say we like those scenarios, but ultimately VMware's got to compete. I want to talk about Oracle, but this is such a good conversation. We'll come back to that. Hopefully we'll have time, I'm sure we will. I want to talk about big data. We've had Val, Bercovici on, and John and I were struck. I think we first had him on the last SNW and he had a very good perspective on big data. I was somewhat surprised. I didn't think NetApp, it was associate NetApp with big data, but since then, and do, what are you seeing in your first role, looking at futures and integration and the like around big data? I saw Oracle through its hat in the ring with a Hadoop appliance this week. What are your thoughts and what's NetApp doing there? So I think it's a very interesting time in the market and obviously I don't focus on big data or even databases in an area of technical expertise, but the one thing that I am certain of is by the time I finish this sentence, there'll be more data generated and being asked to be retained for some long period of time than when I started it. And so scaling is an area that I spend a lot of time and thought around whether it's in the context of the virtual infrastructure or in terms of scaling, in terms of big data, storing analytics of those types of working sets and things of that nature. It's a massive challenge and I would suggest to you that when I engage with customers, I think a lot of folks are underestimating the amount of data that they're going to be required to store and service five, 10 years from now. So from a NetApp perspective, our focus is to continue to innovate so that we can support new constructs like Hadoop or no SQL databases. So the customers that are ready to move from their traditional models or business applications into these new constructs that are rapidly maturing right in front of our eyes can do so. So where this is going to flesh out? I don't know. But outside of the ability to scale to, well, every customer has the same goal, right? I want to do more with less. I don't want to do it faster than I did yesterday. How do we get there? I don't know. All right, let's talk about Oracle a little bit. One of the other things Tom Georgin said, and these are my words, not his, but roughly it translates into the Oracle opportunity. Oracle hates us less than they hate everybody else. So that's an opportunity for us. That's probably true. I think I've said on theCUBE that EMC, in fact, has probably the most to lose from Exadata. IDC released some data today, and EMC is number one in Oracle environments. NetApp was actually number two. But in different segments, I mean EMC, VMAX, high-end OLTP, that's their roots. And Exadata's going right after them. There's no question about it. NetApp has an opportunity, EMC, by the way, has a similar opportunity in data protection. Is that manifesting itself into real business, whether in the channel or in your direct sales force, and in the Oracle customer base? Can you talk about that a little bit? I would say it's absolutely a reality. Customers have been asking for more maturity in the solutions around virtualizing Oracle. The areas that they are looking for this maturity tend to fall in a couple of buckets. Obviously, performance front center. What's the tax for the gains I receive by virtualizing the infrastructure? What's the tax I then pay on the performance layer? So the second area, and I can go into these in better depth, but the first area would be performance. The second area is really around the mature maturation of the manageability tools, whether that's from looking at applications that, let's say, in the backup space, business continuance space, or probably more relevant of late, is in the service level space. How do I gain assurances that my application is running optimally when I've got to look at an application layer, a database layout, hypervisor layer, networking storage, at the physical layer, right? I cross this bridge between application and virtual infrastructure to physical. So I think the maturation of the tool sets in that space allow us to have the assurances. And then the third area is customers, I think sometimes look at the virtualization layer as a intermediary step. So let's say I'm heavily vested in Oracle today. I want to start leveraging some of the cost savings that are available to me through, say, VMware vSphere. Today I want to move Oracle into vSphere. I want to have all these assurances that I can do successfully, but I also want to be sure that after that migration is completed, can I leverage that same infrastructure and components should I need to make a change in terms of how I'm servicing that application, right? I need flexibility, I need to have protection to make sure that I'm not taking on technology dead over the years, right, at the infrastructure layer. And so we're seeing customers really start to state that flexibility that's inherent in the platform at a hardware layer and software layer is a top priority in terms of their decision criteria when they're going to make a purchase. And I think NetApp has played well in that space and I think that is a big opportunity for us to continue to drive forward with it, which is, let me assure that you can move it successfully today in two years from now when some new innovation hits from VMware or Oracle, et cetera, that you can take advantage of that without having to rip and replace your hardware. Now, Von, I realize you're not a deep Oracle practitioner, but I want to talk a little bit about virtualization of the Oracle space in VMware specifically. I mean, it's no secret that Oracle doesn't want VMware installed running its applications in it, you know, will make customers kick and scream before they go there. Nonetheless, customers are going. And they kick and scream and they get support and they get, you know, eventually what they want, there's some, you know, considerations around licensing and the enterprise level. Yeah, right, if they pay, they get support. If they don't pay the support, they get, go back to physical and then call us. But so, but yes, if they pay, they get great support from Oracle. There's no question about it. Or if they do rack in VMware, which, you know, okay, great. But I want to understand what you're seeing there, specifically with regard to Oracle's virtualization strategy. And where does that fit on the priority list for you? I mean, it's been slow, it's been immature, but I understand there's a new release out that's looking more promising. What are you seeing in the Oracle base? So, I think we're seeing two trends. I think customers, when they look at their virtualization layer, are more and more looking at it as a management layer. It's a means where I can standardize and tool sets, operational processes. That leads to repeatability in the reduction of operational costs. So I think it's fair to say that a significant number of business leaders, say C-level, are looking for standardization within their operations. Conversely, I think the database administrators and folks who their day-to-day tasks live in that world are probably much more comfortable with the one-throat-to-choke approach, right? And are worried about the intimate details that affect them, and I think probably are more comfortable in looking at Oracle's virtualization space. And I would also suggest to you that they're probably not as abreast of all the nuanced differences between a VMware virtualized platform and Oracle. So from our perspective, it's volume-wise today, and again, this is always subject to market changes. The request for Oracle and VMware is dominant, but those who want an Oracle stack is passionate. What's interesting is we only get brought into that Oracle stack when the customers say, I need something that NetApp is offering in the data management space to help me with this. Okay, so that answers my next question. Why even bother if they're going to go with OVM? They're probably going to go with some kind of sun disk drive and why do they need you? The answer is because they love you and appreciate what you're doing. It's funny how pervasive a concept can be. So when you think about server virtualization, everyone understands that concept really well. If I say server virtualization in your mind, you think VM, hypervisor, I don't need to say anymore. If I say storage virtualization, we all get, well, that's a gray area. Many, many shades of gray. We're not clear what that means. So one of the things that has worked very well for us and why we're pulled into these conversations around virtualizing Oracle is we have a unique ability to change the IO pattern within these data sets by being able to share blocks of data that are cached in the array across multiple database instances, and that's truly unique in a storage, from a storage platform, historically a storage array, whatever is cached from one host is only available to be serviced to request of that host or subsequent requests from that host, and we break that IO paradigm with some of our capabilities, allowing customers to scale further without having to necessarily have a direct correlation to the hardware that they're purchasing and where that really has an impact is driving down cost in terms of test and development and the lab type of scenarios. We've always been very strong. It's also allowing us to have more opportunities as we go up-stack into the larger databases, customers who historically maybe have been on a symmetric platform that are looking to say, how can your platform best serve my workload? We kind of come out of an approach that's different than the classic storage tiering and storage caching, and we start talking about, well, IO, right? Do you want to be able to scale your IO at the hardware level or do you want to put some intelligence in this and scale further? And I think that's worked out very well for us. Good. We're here with Vaughn Stewart, who's with NetApp. Great discussion, Vaughn. I really appreciate you allowing me a lot of latitude. A lot of latitude on my side. Sometimes I digress a bit. That was just good. Really would love to have you back on. NetApp, you know, doing a great job in the marketplace, once a small company, now a big company, the last big independent storage company out there. I mean, EMC is really not just a pure place storage company anymore. You guys are. And so congratulations on your success. Appreciate that. And thanks for coming on theCUBE. Thanks so much. I appreciate it. Great to see you. Take care, Dave.