 Live from San Francisco, extracting the signal from the noise, it's theCUBE, covering Oracle OpenWorld 2015 from Studio C, brought to you by Cisco. Now your hosts, Stu Miniman and Brian Gracely. Welcome back to theCUBE, SiliconANGLE Media's live production of Oracle OpenWorld 2015, here on the exhibit floor in Moscone South. I'm Stu Miniman, joined with me from my co-host. On this segment is Brian Gracely, and we're both from wikibon.com. Happy to dig in more with some of the stack solutions that are being built on this segment. Ajay Singh is the VP of Product Management with Nimble Storage. Ajay, welcome to the program. Thank you, great to be here. All right, so Ajay, you're meeting with customers, you're talking with them on lots of various topics. You know, what does kind of the solution set that Nimble's a part of look like to your customers today? So what we hear from customers a lot, of course, part of what we're all about is flash-based storage. And we're right in the midst of this giant transition from legacy, disk-based array architectures to a newer generation of architectures that are designed to better leverage flash. And the reason, obviously, being a much better mix of price performance, cost of capacity, cost of performance. So Nimble's been one of the new generation leaders leading the charge on that. So we hear a lot of requirements from customers, particularly related to databases, for example. Because databases are typically the foundation of mission-critical applications, need lots of performance, needed cost-effective cost of capacity, ease of scaling, and so forth. So we hear a lot about that. Since they're also mission-critical applications, we hear a lot about customers wanting proven, validated solutions, converged architectures. And this is where Cisco has been one of our, in fact, the biggest co-departner in that regard. So over the years, we've done a number of what we call smart stacks, validated architectures, specifically purpose-built for the application the organization has in mind. One of the biggest ones we did was JD Edwards running Oracle on UCS blades on Nimble. So what the customer's getting in that is fully tested, purpose-built, appropriately sized, and jointly supported solution for a mission-critical application. And we've been one of the newer generation vendors. I have to say we didn't create the trend, but we are really pushing that trend forward in partnership with Cisco. Excellent, so we love that. If you could simplify the solution, that helps customers. And sometimes it's about doing it better, faster, cheaper, easier than I did before. But sometimes it's also about what can I do that I just didn't have the capability before? Either freeing up resources to do something different, or the architecture just allow me to just do new things, either with storage and leveraging that, or just innovative new solutions. Any specifics you might be able to share? Yes, in particular, I would highlight two things. So one, no matter how well you size your initial deployment for a converged infrastructure, IT needs scale in unpredictable ways. Maybe you acquire a new organization. Maybe you add a new project to your plate. And the requirements for performance capacity suddenly go beyond the scope of what you originally envisioned. And this is one of the areas Nimble shines. We have unparalleled capabilities to scale performance, capacity, every resource within the system, independent of each other. So we call this sort of a Lego block approach by just what you need on day one. Knowing you have no risk of running into a limit, you'll be able to add just the resource you want when you want, whether it's scaling capacity, performance, or clustering horizontally, and thereby assuring your investment's well protected. So that's one thing we do really well. The other one we do extremely well is what we call operational analytics. It's an offering we call InfoSight. We collect tens of millions of counters every day from every one of our systems, and we have thousands and thousands around the world. Analyze them real time with two goals in mind. Support should be proactive and prescriptive. We tell you before you hit an issue, you're about to run into something, and here's what to do to avoid it. So customers love that. It feels like we're an extended part of their IT team. The second value is we sort of open up the black box. IT hardware is often a black box. You have to hire professional services to know what's going on inside, and we give you this rich window into everything. Usage, growth patterns, troubleshooting, everything about the infrastructure, and we've been extending that up the stack to do that not just for storage. We are now able to do that for VMs, and in partnership with Cisco, there are other interesting avenues that have been extended into. Yeah, so I love that. There's really two pieces you talked there. One, Wikibon's done studies of just productivity of the typical employee, and we've said it's not just about getting more out of the people, but the true business value that we've had there. And secondly, it's getting more into your data. Leveraging that, I guess the question I have, do you have ways that you're helping customers manage that, really get visibility in what they're doing? I know that you're just interested in things with your InfoSight software, so how do you kind of make IT the new hero? So that's one part of it, what I just described. InfoSight gives you a level of visibility that they've never had for 30 years of managing storage. The other part is to do with our usability and management tools, and what we're doing to integrate them up and down the stack. So of course we have our own user interface, almost any storage architecture that's built in the last five years. The bar should be getting up and running in the 30 minutes, not three days or five days it used to traditionally take. So that part's table stakes. What we do is integrate it with the entire stack management tool. So with Cisco, for example, we have plugins for UCS Director. So now you have a way to provision, monitor, and manage the entire stack, not just the individual layers of the stack. So you're bringing, you're simplifying how I can acquire the infrastructure, how I can use the infrastructure, and then you're bringing sort of a 10x, an order of magnitude of speed that's faster. Great underlying technology. What does that translate into for your customers' businesses? When their business owners think about I got a project, what's their change in thinking? So there are two classic scenarios. One is where the business has a problem. They have an analytics application or a database or queries that instead of running at eight hours at a shop, they want it to go much, much faster because that eight hours is slowing down their pace of development. It's slowing down their critical business process. It's costing them money. It's costing them opportunity costs. And we're often able to bring that eight-hour, 10-hour, one-day query or one-day batch process to minutes. So that right there translates to business value. You're answering business questions faster. Yeah, so unlock a project barrier that's slowing down the IT organization or maybe preventing them from jumping on to more mission-critical projects. So that's one class. The other class is more mundane. They don't necessarily have a huge, crushing business barrier, but it's the capex and the cost of managing this infrastructure that's killing them. Their three, four-year-old storage system that was one of the big vendors is coming out for a support renewal. And they've heard enough about the market to know there's better approaches now. And they want to save, they want to bring that three, five racks of gear down to half a rack if they can, dramatically simplify what it takes to manage it and of course, save a lot of capital in the process. So customers typically fall in one of those two categories and we are really good at solving both. Yeah, as you're going faster, delivering faster, deploying faster, how much do you guys feel like you have to compare against the sort of frictionless of public cloud? Lots of cloud talk this week. Are your customers asking you for experiences against some baseline? They do, particularly because there's such an awareness, it hired management layers of IT about what should be possible, swipe a credit card and you're up and running, working on or developing a new application, versus spending weeks planning your next infrastructure expansion. So one of the ways we can prove our business value is one of our fastest growing customer segments is infrastructure as service providers, so cloud service providers. So they're not quite Amazon or Azure, but then there's a tier of service providers that provide higher level of SLAs, more control, more customized services, better control over sovereignty. More enterprise cloud applications, specific clouds, yeah. And that's where we're able to prove our value by helping them deliver services to the enterprise customers that are very quick to procure, install, configure, and then very quick to provision every time a new admin needs a new application or a new bunch of VMs to deploy, we're able to match that experience that they would get from a public cloud. Yeah, so Ajay, it's interesting, I was hearing you talk about kind of the C suite pushing some of the cloud discussion, or kind of the IT people really looking still at kind of the individual components and where do some of the system discussions, you talk about the stack solution that you put together with Cisco, I know when V blocks first got out, there were same discussions, it's like, oh, that's a C suite discussion, not a IT administrator discussion. So the C suite is typically looking at that what they call time to value, they want to know they have a critical business initiative to deploy a new application, to design it, to acquire all the infrastructure they need to get it up and running, and then to ensure that it's going to succeed instead of a project that's going to fail six months down the line because of unforeseen barriers. So the role Nimble and Cisco, and often a third application partner play in that is delivering a pre validated architecture that's purpose built for that particular application. VDI was a great example over the last three or four years there's been a wave of organizations adopting VDI for the first time, but we've also seen a lot of organizations doing it for the second time because the first one was a disaster. And where our smart stacks have played a great role is assuring a higher level of success and doing it quickly. You can acquire it faster, it's completely pre designed for you, you know exactly what skew to acquire, and you know when you get it up and running, it's going to work, it's going to deliver as promised. The CIO loves that because what gets CIO's nightmares is the risk that a mission critical application won't come online as planned. Our David Floyer, our CTO Wikibon has written a bunch of research that really looks at, you know is it feasible to do an all flash data center? And his conclusion over the next five, six years is the economics are going to work out that that's possible. How many, you know, we're seeing duplication, we're seeing compression, you know, a lot of things in the technology that start to bring the cost down. How many of your customers are starting to push you already to say, hey, you know, other than maybe a very low tier of disk, you know, kind of an archival, want to push to all flash? We do get that question. It's not so much what percentage of customers, but rather what percentage of applications within the typical large data center. So we've adapted our products along the way. It used to be very, very fixed ratios of flash to disk. A couple of years ago, we allowed a much richer range of flash to disk ratios. Last year we introduced something called the adaptive flash platform that lets you get 100% flash to disk ratio if you want. And recently, earlier this year, we introduced an all flash service level in our systems. So we can go from an all disk service level for the maybe the scratch space, the test and dev applications that don't need a lot of performance to a auto flash service level, which is sort of our bread and butter for the last five years. And a more recently introduced all flash service levels that lets you guarantee sub millisecond latency for every single transaction for applications that need it. And we've stated publicly, our vision for our flash portfolio doesn't end there. We do agree that there's going to be an increasing need for more applications or an opportunity for more applications to take advantage of richer flash architectures. And we fully intend to participate in that. So Ajay, I'm curious as you're talking to customers, how much of the listeners say Oracle who says, well, just buy it all from us, from the silicon all the way to the application. We're looking at Microsoft, taking Office 365 and putting the applications in the Azure cloud. How do you compete against that and what's the conversation you're having with customers today on this topic? I do encounter both and the conversation typically is, do you think you can offer something better and do you have proof points? So I'll take Office 365 as an example for a medieval competitor because they own the whole stack and have many levels of flexibility in how they die, licensing and other things, plus an inherent advantage in customers thinking, well, they must know how to do it and they can probably do it cheaper. And I have to say partnering with service providers who specialize in Microsoft applications, we've been able to take on cloud service providers, including Office 365 for perhaps the toughest example is Exchange. A lot of people think if Gmail can be done in the cloud, Exchange can also be done in the cloud. And partnering with a couple of big service providers, we have now one time after time, a Fortune 1000 customer who has moved tens of thousands of mailboxes to a stack where Nimble is the core. And the reason we've been able to do that is proven capital and operational efficiency. So we are able to save a lot more money, provide a lot less hardware to deliver the same performance and capacity that an Exchange deployment would need and to be able to scale it as you go from maybe 20, 30,000 to 100,000 employees. So we have lots of proof points in ensuring we can compete successfully in that space. The short answer is a fully owned stack has a perception advantage to begin with, but if you can show that you truly have actual tangible advantages, then you can convince the customer to break open the stack and buy best of breed. You know, I like that, Joe. I just think about, you know, let's take Exchange for example, you mentioned, it's boy, you know, the whole storage industries live through pain for the last, you know, 15 years deploying some of these. And to be honest, how much value is there for any one company to like be an expert at deploying their own mail? I mean, mail's critical, don't get me wrong, but I like the service provider angle. You're giving customers some flexibility as to how they deploy things. Certainly that's been our bet. Yeah, you know, one of the big themes from the database perspective here at Oracle OpenWorld is more and more moving into in-memory. You know, and, you know, we talked to the, some of the Excel block stack people and they were saying, well, we're seeing more and more in-memory, which means less reads. What is, what's the future of storage arrays when we're seeing more databases? Is there still a place for it? Is there value add you can wrap around in-memory databases? Yeah, there is. So for example, we have a number of certified architectures with SAP HANA. That was one of the first to pioneer and really lead the charge on in-memory databases. So you're right, the IO characteristics are different, but the persistent storage is still the shared SAM. So you still want it to be able to handle lots of IO. It's a little different in characteristic instead of lots of small reads and writes. You might have big bursts of reads and writes. And importantly, you also want it to provide characteristics that a mission critical application needs. Rollbacks using point-in-time snapshots. Very efficient disaster recovery. And for the test and dev users, really, really efficient zero copy clones. So there is still a big role for shared storage, in other words, SAMs to play in that environment. It just looks a little bit different than it does with the classic transactional databases. Got you different IO profile that you had before, got it. So Ajay, as you're at the show here, what do you look forward? What's different about Oracle Open World today compared to some of the other shows that you go to? What I look for is what's here this year that wasn't here last year? And usually those are things that are found around the edges because the bigger vendors typically don't tend to change from year to year. We do see a lot of vendors promising to solve the performance problem because often there are opportunities to save on database licensing. If you can ensure the CPUs aren't waiting for IO at the storage layer, so there's a number of vendors in storage and outside promising to solve that problem and we certainly play a big role there. We see a lot of vendors talking about it and analytics, both in understanding customers' data but also understanding what's going on inside your IT infrastructure. And again, that's where InfoSight is. I would argue the leading approach in our space, certainly. And we're starting to see interest and recognition that there is a category of solutions that provide that. All right, well, Ajay Singh, Nipple Storage, really appreciate you coming by, sharing with us, everything go on here. Please check out siliconangle.tv. Not only does it have all of our live events, lists our upcoming events, and if you see something that we're not currently at that we should be, click the button at the top that says bring theCUBE to your event. Thanks for joining us. We'll be right back with lots more coverage here from Oracle Open World 2015.