 Extracting the signal from the noise, it's theCUBE, covering VMworld 2015. Brought to you by VMware and its ecosystem sponsors. Now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are live in San Francisco at Musconi North for VMworld 2015. This is theCUBE, our flagship program. We go out to the event, and extract the system of noise. I'm John Furrier, my co-host Dave Vellante. Next guest is CUBE alum, Craig Nunez, VP at HP Storage. Welcome back to theCUBE, great to see you. Thanks, great to be here. As always, my favorite part of VMworld. Sixth year at VMworld. Nice. We were talking to Vish earlier. Three part the time was an independent company, and it was at VMworld, when we just sort of documented the bidding war. Yeah, right. Between Dell and HP. And we predicted accurately HP would win. That wasn't a hard to picture. I can only remember the compelling them was on the scene, all that stuff's happening. Craig, what's changed? I mean, obviously the tweets have gone out. We put out a tweet as well. It's just, it's like a storage shows. The joke is, it's storage world. Absolutely, what's going on, why? Well, I think we are at a crossroads. The industry, so the industry is always moving, right? We love that. But there's a tectonic shift going on right now with all that's changing, with flash washing through the data center. We have a really interesting development with hyper-converged coming through for folks looking for that kind of agility. And the bottom line is, with so many new kinds of applications coming through, folks trying to power that digital economy, they're having to kind of stop buying the good enough storage that they may be been buying for the last 10 years and kind of look around and figure it out. And I think folks are rushing to kind of fill those new needs, those new gaps. And that's what makes our industry fun. And the DevOps thing is also powering this change. Absolutely. The application workloads are shifting the requirements on storage. And what specifically are you seeing there? I mean, a lot of it is around the time to value capability that you bring into the data services, the storage itself. And traditionally for folks who've been involved in storage for years and years, not the strong suit of the storage infrastructure, agility was not really in the same sentence. And so part of what's gone on, I think, starting with 3PAR, we really built a platform for that virtualized data center bringing the agility that we're observing on the server side, carrying that forward to really drive that kind of time to value for, in fact, the old days, the DevOps guys working at the service providers doing cloud-based hosting. So bringing them away to shrink book to bill. Hey, I've got to fill this customer deployment. What's the fastest way to get it there, 3PAR? And that's very much a theme now that a DevOps world has gone mainstream. So 3PAR, I always tell you, I still have the original 3PAR S1 on my desk, all highlighted, you know, you pour through those things. And I mean, there was obviously a lot of great engineering, but there was some luck involved there because the theme at the time was all about service providers, you know, SNI, you remember storage networks, right? And it was crazy that that was going to take over the world and the S1 read like, hey, this is utility storage, essentially it didn't say it, but we are building this for the cloud. And that's what you did. And it just was perfect timing. The vision back then, which holds today was, IT is going to become a utility and only those with scale, whether it's service provider or an advanced enterprise who's really got that kind of scale is really going to be at the end of the day, got the scale to power at the lowest possible cost. And you know, folks are in various stages of moving there and we're seeing, you know, adoption of these internal utilities, you know, called them private clouds. Hybrid clouds today is the vernacular. We had different language a few years ago. But I mean, you couldn't have nailed it better. And so you addressed a lot of the complexity problem. You know, we used to call it the click wars and three power was the gold standard of that. You really simplified, you know, the whole setup deployment installation, obviously rich set of services. But now we're entering this sort of new phase and a flash is changing everything. You got cloud really happening now. It wasn't just sort of a bunch of marketing BS. And so the pressure for agility is even greater. So how does an organization like HP with an architecture that's built in the 2000s granted for the right market segment, how do you keep pace? So here's the way I think about it. Cause first of all, there are smart people everywhere, smart people at HP, at EMC, at NetApp. I mean, they're in our industry is filled with smart people. They, what kind of teases vendors apart is the architectural assets that they're bringing to market and those assets enable those smart people to do some wonderful, wonderful things. And by the way, I didn't come on today to talk about three part, but I'm happy to do it because that platform was built to drive massive server consolidation. And in those days, it was these big huge, you know, blade frame kind of devices and VMware was still sitting on people's workstations. That of course moved to ESX and, you know, the world as we know it today. And that is even looking like it's gonna move to a containerized kind of workload. So you can see that the trend, a platform built for massive server consolidation keeping pace with kind of the trend on the front end. The back end built for, you know, massive parallelized workloads to try to get all you can out of every single drive, every IOP and point it to a specific VM, right? And a system built for that actually is a system built for Flash as well. Cause think of what Flash is. Flash is just a massive ton of IOPs that you wanna aim at a specific workload, right? And so, you know, call it luck, call it I think foresight by some of our smart architects at three par. Yeah, that platform has really, you know, allowed HP to take a new found position in the market. Number one in mid-range fiber channel. Number two in all Flash. And pretty darn close to probably number two overall storage supplier in the industry. So not a bad position. So you guys talking about these other trends, you talking about DevOps, you know, you got the sort of open source, open stack trend. Yeah. You mentioned containers. Containers in particular is, you know, that's a pretty big trend going on. A lot of momentum there for Docker in particular, but others, is there an affinity? What does it mean for storage? You know, that whole containerization, virtualization meant a lot. That means a lot. This is Johnson still a storage show. What does the container trend mean to storage? Will there be a greater affinity between the data repository and containers? I think, you know, we've got to see how it plays out because I think, like VMware in the early days, there's an evolution ahead for, you know, containerized applications. I think, simply put, storage is, you know, faced with yet another set of unpredictable workloads. What are those going to look like? I just figured out, you know, VMs and SQL server, right? What is a container workload going to look like? And the platform simply has to be able to respond without somebody at the keyboard. Otherwise, the whole business is going to be mired down in, you know, the, you know, kind of behind IT. And that's the last thing we want. So the platform really has to handle those unpredictable workloads and deliver a predictable quality of service to every single app, you know. Yeah, it's okay. So if you're going to go from, you know, a handful to tens of thousands, maybe even 100,000 containers, you got to keep the response time per flat. And so the infrastructure to do that has, I mean, I would think has the change from where we are today. Totally template driven. In fact, you know, we thought a lot about this when we developed our, you know, kind of our VVOL support, because that too was kind of going to raise by an order of magnitude, the, you know, call it the clients, the applications that you're supporting. And so everything has got to be automated, template-based. Otherwise, you'll never get ahead of the wave of work coming at the, you know. What's your take on the show? So drill down on your perspective on the storage aspect. Why is it so hot? I want to revisit that because, and I want to, I don't get what HP's doing because this is a big part of the choice message we're hearing. The VMware ecosystem is not a homogeneous environment anymore. Certain layers VMware has, we heard from Pat Gelsinger. Yeah, VMotion, stuff like that. But outside of that, it's a heterogeneous world. HP's always been multi-vendor. But what are you guys doing here in context of the ecosystem? Because it's a free-for-all right now. VMware, it's a lot of growth, a lot of install base, a lot of storage needs. What particular things are you guys innovating on? So the, I think the big things that were, you know, very much in call it education mode with our customers and folks who aren't our customers. Number one is Flash, because I think we have seen this, you know, Flash innovation and the premise I hold really came true with Flash. Great innovations, you know, really get kind of re-innovated a couple of times before they go mainstream. And Flash was, you know, awesome for those app acceleration, revenue-tied applications, but very niched out, right? And folks started to think, hey, I could bring this into my infrastructure if only I could bring, you know, greater tier one capable data services. It would fit, I can do this. And so we saw Flash kind of grow in that direction. And then the, I think the one that really, really tipped it over was, that's great, but most of my apps are sitting on disk and that's, you know, a cost per usable gig of X. And when Flash hit that, you know, buck 50, $2 per usable gig, bam. We have folks driving towards this all Flash data center. So kind of that tipping point was cost per, cost per usable gigabyte, which a lot of folks still think Flash is too expensive. Either they're talking to the wrong vendor or, you know, they're kind of, you know, last time they looked, it was too pricey. So that, getting that education out, the other, you know, hot topic and you walk around the show floor, it's evident is hyperconvergence. I think everybody is talking about it, whether they've got it or not, it's a different thing. And it's where Flash was two years ago, right? What's hot right now is an appliance, right? A fixed building block for VMs, client desktops, whatever, but we know a great innovation is going to reinvent itself a couple of times. And what we're beginning to see when we talk to some of the more thoughtful enterprises is they want those data services captured within the appliance to stretch beyond the appliance, right? And be a part of their data center and help them solve problems like low cost replication, maybe with software on an x86 server or, you know, software running on bare metal hardware to just bring additional storage capacity to that hyperconverged appliance, to break out of that fixed building block mode. And that is going to, you know, be like the tier one data services. It's going to increase the adoption. And I believe what drives that stuff mainstream is when those data services associated with hyperconverged infrastructure really become a ubiquitous data services across the data center with all that cool flash that's, you know, powering your most important applications. When you have one set of data services across hyperconverged and flash-based systems, bam, you have a rational, coherent data center architecture that you can bet your business on. So that sounds like a renaissance in metadata management that occurs to make that happen. Absolutely. And it's, we all have to kind of step back and not solve the point problem first, but, you know, solve it in the context of where we're taking our data center, right? Because if we're kind of caught up solving point problems, we're going to wind up with silo after silo of things that don't work together. Well, presumably that's a problem that a company like HP and the others that are large have the resources to solve because generally startups are solving those point problems or is that changing? I mean, what's your thought? Well, I think in general terms, hyperconvergence is a great one because that is a problem not just addressed by great data services. It's, you know, it needs great management. It needs great dense hardware platforms. And if you think about a company like HP that brings a fantastic compute portfolio, it brings one view on the management side. It brings great storage technology. You know, there's no other company I can think of that really is bringing it together like HP and can really take it to where, you know, it's capable of going. Craig, great to see you. Thanks for coming on theCUBE and sharing your perspective. We've been here six years, as Dave said, and having you on theCUBE's been fantastic for always great insight. And again, Dave, storage is sexy. Again, every year, always, babies. Bring it. Storage is sexy. And again, this show is dominated by storage because there's so much going on. The capacity issues, the technology, the need for the big day that I'll see with virtualization, a lot of great stuff. Congratulations on all your success at HP Storage. We'll be back with more live in San Francisco from theCUBE here at VMworld 2015 after this short break.