 From London, England, extracting the signal from the noise, it's theCUBE, Covered, Discover 2015, brought to you by Hewlett-Packard Enterprise. Now your hosts, John Furrier and Dave Vellante. Hey everyone, we are here live in London, England for HP Enterprise, HPE Discover 2015. I'm John Furrier, the founder of SiliconANGLE, and I'm at Coast Dave Vellante, founder of Wikibon.com. This is theCUBE, our flagship program. We go out to the events and extract the signal from the noise. Our next guest is Bill Bin, VP of Virtual Development Unit Storage at HP Enterprise Storage. Welcome back to theCUBE, great to see you. Hey, John, Dave, how are you? So storage is still hot. Again, since 2010, it's been the hottest thing, but it's changing more now than ever. You're seeing a lot of companies when public peer storage went public. That's right. And kind of like flatten out, you're seeing a lot of M&A activity. NetApp's trying to figure things out. Software is now shifting the game, got all flash arrays, you have all different kinds of architectures. Customers are looking for new ways to take the storage and go to the next level, the next big thing. So what's going on with you guys? Take us through some of the current news and how does this relate to those building blocks of storage? Well, I think the storage business is radically going to change over the course of the next couple of years, right? And I don't know if it's just the new entrants in the market or if it's the Dell EMC sort of, if that deal sort of actually gets consummated or not, but it's clearly the storage industry is changing. One of the things that we see changing the most, obviously, is the customer's decision desire to move storage closer to the application. So whether it is the pure software defined, whether it is the hyper-converged plays that are out there, or whether it's frankly just being able to consume the infrastructure they have, all of the composable infrastructure announcements we've done this morning, it's all about customer choice, right? And I think what you're seeing is we're moving from these sort of converged systems which are fixed sort of building blocks and the hyper-converged guys suffer from the same restrictions to a composable infrastructure which is all about customer choice, right? I want this much compute and this much storage. If you do that in a traditional hyper-converged box, you have fixed compute and you have fixed storage requirements. And if you don't like that or if you need more, you buy another box. But HP, we actually believe the customer needs choice. So you want to buy a hyper-converged box, we got one of those. You want to do software defined on a server, we got one of those. You want to do conventional storage with three-par, we got one of those. And the benefit of that is it's not sort of the least common denominator, right? These are all best-of-breed solutions from between store virtual and store serve, which is three-par, best-of-breed, all sort of come provisioned in the same common user interface. So we think that there's, you know, that HP is actually well positioned for this, for the environment. The number one question I get asked from practitioners when I hang out with the Wikibon analysts is, hey, I love all the speeds and feeds and stuff going on in the industry and I love more lower latency, all that stuff. I want more of everything that's coming around the corner. But I got an engine room called Storage and Compute and Data Center that I don't want to screw up. And I want to actually evolve it. So the best-of-breed thing kind of comes out. So take us through some of the things that you're here. And the number one question is, how do I get the future without killing the present, per se, because they are running an engine room per se. Storage is power in their business. They're kind of adding some innovation around it, but they want to get to the next level. How do you talk to that customer and what does HP have for that? Well, I think what customers are really looking for, first of all, is simplicity, which is why I think there's some of the attractiveness around the composable infrastructure story or the hyper-converged story. It's all around simplicity, right? And that simplicity oftentimes is relative to the application provider or the provisioner. One of the trends that we're seeing in the industry is, you're moving away from sort of these purpose-built positions, a storage administrator, a backup administrator, et cetera, to actually enabling the application administrator to actually own the provisioning of storage or the provisioning of data protection. Because it's all around cost ownership at the end of the day, right? So one of the things that, if you look at sort of what HP has done with the HP OneView announcements we've done today, is that you can actually provision server storage and networking, the same common framework, which is the one new framework, right? And it actually makes that application enablement much more effective. And I think that, to some degree, that's where the hyper-converged guys are sort of leading the business. The last time you and I talked, we were talking about cloud, if I remember. We were talking about that dirty little secret, I think I said, which was behind all those cloud implementations is actually conventional storage, right? I think the same is essentially true with the hyper-converged vendors in the sense that while cloud is attractive from easy to provision, the cost, et cetera, people still want enterprise reliability, resiliency, et cetera. I think the hyper-converged vendors are also going to drive us the same way, like our HC250, which is you can build it, install it, and get it running storage in less than 15 minutes. That sort of easy to use application deployable technology is where the whole industry will move to, but eventually it's going to move back to the guys who know how to do enterprise storage. Yeah, and since we last talked to, the cloud has also changed the mindset of some of the practitioners. They see cloud as a great way to say, okay, we're DevOps now, which is agile, which is basically code word for, hey, I need more storage now, so the scale question comes up. So it's not so much doing cloud per se, like doing some cloud is easy, relatively easy. Scaling, it is very difficult. Well, it's the 80-20 problem, right? The first 80%, you spend 20% of the work to do, the other 20%, you spend the rest, the other 80% actually get it done, right? But I think that... How do you guys address the scale question? Because that's really the thing, like, okay, I'm bought into the cloud and hybrid, I'm going to do a little bit of public, mostly on-prem, and some hybrid, but I got to scale up now. I got to really build out my, and spend some more cash, and get my app guys programming with low latency, in memory, all that good stuff. Yeah, and I think that's, I think if you take a look at work we're doing around OpenStack, or the initiative that HP Enterprise now, I have to say, is doing around the Docker and other container initiatives. It's actually actually making storage really easy to provision behind that prevailing technology, right? So a lot of good things happen in there. You're off for quiet, I was going to say. I'm going to jump in now. So we talked about composable infrastructure before, and sort of juxtaposing it to the traditional, now traditional, the hyper-converged guys. If I understood what you're saying correctly, you're able to scale resources independently. Correct. And so I wonder if, and that's really the main difference. That's right. I wonder if you could talk about that difference, and then specifically how you maintain balance, right? Cause the hyper-converged guys would say, well, you know, we could kind of do that too, and we're going to maintain balance. So I wonder how you respond to that, and how you guys approach that. So the way that we think about the problem, Dave, is that unfortunately applications don't simply behave with fixed compute and fixed storage boundaries, right? So I don't have this nice and easy relationship between the two. So you've got to be able to do a couple of things. One is when you're out of balance, or you've got an application that's heavy storage, or light compute, you've got to be able to extend off the back of a hyper-converged system, right? And so one of the other interesting things that we see happening is these hyper-converged systems are like any other bespoke appliance in the data center. Over time, like the all flash systems that used to be out there, over time customers would look to actually move those to a conventional storage or simplification effort, right? Hyper-converged is the same way. So one, you've got to be able to extend storage off the back. Second, it's got to be able to talk to what you've got within your data center. So for example, store virtual, you can deploy a physical appliance, you can deploy a hyper-converged appliance, or you can deploy a software-only appliance, and the data federates, and you can actually peer motion between them. Okay, so in that way, customers actually, in the case where they've got a hot application that's running a hyper-converged platform, you can actually peer motion it to a physical appliance. I think one day, what you should be able to do is actually be able to peer motion that, not only from a store virtual box, peer motion it actually to a three-part box, right? And that's all about the choice that we're talking about. Okay, so I wanted to ask you about the portfolio, and you guys have made a lot of effort to try to simplify the portfolio, and at the same time, do some base-level integration. Peer motion is an example of that. Talk more about that sort of integration strategy. So if you look now, you look now really what we're on a path on is you've got one view for sort of the composable infrastructure story. Now you've got store virtual and store once as well as store serve, right, three-part. Actually now employing the same sort of user interface that the one view guys have used. So now it looks like all these things come from the same company. More importantly, the training effort between the two is greatly reduced. So that's the first answer. The second answer is within that, what you ought to be able to do is actually be able to not just online migrate, which is really, it's a one-way transaction, which we can do today. You ought to actually be able to federate not only homogeneously between three-part boxes between store virtual boxes, but you also should be able to federate from a store virtual box to a three-part box. Now we're not quite there yet, but that's essentially where we see the mission going. So if you've got a remote office site, where you've deployed a hyper-converge and AC250 platform, you've got three-part in your data center. Data should federate between the two. So that not only for a disaster recovery scenario, but also for a work balance scenario, you should be able to do that. And actually, it should be a non-event from a customer to actually know the storage that's sitting underneath. That's sort of where we're headed. Okay, so you're putting a lot of your time and energy into simplification, I'm hearing. I wonder if you could talk about how that is rippling through to the channel? Yeah. Because most of the channel today still is selling tin. Correct. But they all know they have to transform into more of a sudden selling. Well, not their customers will, you know, the customers will move from them, right? Yeah, right, they're just going to disappear. Right. And so, how were you supporting that? I mean, how does the simplification initiative sort of support the channel? What are you seeing there in terms of what they're asking for as an extension of the customer? So, you know, if you look at what customers are looking for, is they're looking for simplicity. There's still a lot of opportunity for our channel partners to actually add value around it. Today, if you're going to go off and consume a competitive product, let's say, you know, three little word, acronym, you've got to be a specialist in backup in three different versions of FASAN, NAS, et cetera, et cetera. Right? Is the channel partner really adding value there? Probably not, because the customer actually doesn't care to some degree. It's hosting their applications. So through simplicity actually allows the channel partner to actually understand the customer's business better and actually drive a set of integrated services around fewer moving pieces in part. So I think that's actually a good story for a channel partner to get their head around. Yeah, so, ladies, my next question, which is, as you simplify and then enable the channel partner to transform their business more into solutions, how does that affect how you do development? Is there more of a solutions focus? Is that sort of, maybe talk about the roadmap? Absolutely. So I think if you think about sort of where we're headed and I'll call this an application centric model, more and more of our customers are driving their choices they make from storage or data protection from the console that the application is running within. So what they want is simplicity. At the point I provision a VM, they want to actually provision data protection at the same time. I was at a customer in APJA about six months ago. They had provisioned 5,000 new virtual machines. They were very excited about it, right? But they neglected to tell the back of the administrator. They had provisioned 5,000 virtual machines. And they had an outage and they had data loss. If at the time that provisioned the virtual machines of the VMware administrator or the Hyper-V administrator could actually provision storage and data protection together, those kinds of problems wouldn't happen. Another good example is people who want to, database administrators who in a DevOps model, that's the new catchphrase, right? We're out and calling the back of administrator to make a backup to restore to use for a Dev and testing model. With the StoreOnce product, they can actually sit and work with our men and conduct a backup and make a backup on their own. So that whole application-centric administration model, which scares the CIOs, frankly, is where we see the business go. Scares the CIOs because- Now you got more hand. A, you know, you have some determining degree of loss of control. You're making it self-service though, right? Exactly right. But loss of control, single point of glass, if I'm looking at data protection, where I've got everything in my backup ISV, data protector, what have you, those sorts of constraints is a trade-off between application and self-enablement from central control, right? And that's what sort of makes CIOs nervous. It's arcane, but it works, and they have a body on it. Exactly right. But it doesn't scale. It doesn't scale. And it's going to be a disaster to your example before. That's exactly right. How do I back up 5,000 VMs if I don't know I have to back up 5,000 VMs? Exactly right. So I think we'll see that, we'll certainly see that C change. You know, I was with the customer this morning, exact same conversation. That's that C change they're going through, which is where do they actually draw the lines between that central pan of glass or data protection and the fact they want to enable their application administrators to more rapidly respond to business requirements, right? Good news is you can do both with HPE, but we see more of this on the application-centric side happening. And the point you're making about backup is it can't keep being a bolt-on. It's got to be part of the initial design and architecture. Correct. And that's, you know, if you look at software defined, you know, last time I think I was here, I said, and my kids like this, I said we were cool for school before everybody was talking about software defined, right? So a virtual now, almost 10 years inside of Healer Packard was doing software defined from the very get-go. We not only have a software defined story for block, we also have a software defined story for data protection as well, right? Because you can't do one and not have the other. And I think over time what you'll see is, you know, maybe, you know, software defined version for file or, you know, sort of that those steps will continue along the way. So you can actually build, if you will, your own version of the storage, that, you know, storage layer that you want. You know, we've got a customer who takes store virtual and has deployed it on DL380s, you know, from HPE server team, and has put them in all of their, you know, remote office locations because they have the skill set effectively to do that. And it replicates nightly back to the store virtual instance in the data center. And the other part of your R&D story that I'm hearing is choice. Correct. So if you want an appliance, we got an appliance. If you want just software only. Correct. How are you seeing? I actually think what will end up happening, Dave, is that customers won't have to make that, you know, choice also is complication, right? And so why make that, why have to make that choice? So I think what you're going to see over time is rather than deploying a physical appliance versus a virtual appliance, we ought to make deploying the virtual appliance similar to what the look and feel and capabilities are in the physical appliance. So today you apply a virtual appliance and you've got a hardware issue. It's up to you, the integrator, right? The end customer to actually determine what happened. If you deploy a physical appliance, guess what? Something bad happens with hardware, the appliance takes over and actually alerts you and makes all those changes. Why do you have to make those, why do you have to make that choice? I think what you're going to see is our ability to actually bring these two things together. So if a customer wants to deploy a virtual appliance in all of its glory out into their ecosystem that we should be able to enable you to do that from Hewlett Packard. I think that's what's going to end up happening. How about, I wonder if we don't have much time here, but I wonder if you could weigh in on some of the changes that you're seeing in the industry. Dell acquisition of EMC, you saw Jonathan Cure when public, David Scott's famous comment about these guys that they won't achieve escape velocity. One of you could weigh in. They went public with escape velocity, at least from a liquidity standpoint. Pretty good exit, but we'll see. David Scott, who is he? Yeah. Yeah. He's a legend. He's a legend. What was the phrase? Polymorphic. Polymorphic. Polymorphic. Yeah, we no longer use that. Yeah. So, I think a couple things. As I said earlier, I think in the next couple of years, the whole storage climate is absolutely going to change. And I don't think it's driven by cloud. I think it's actually driven by customers wanting different options. That's number one. Number two, there are guys out there like Pure, competing against 3-par. I think we have a really, very good story with them against 3-par. You have EMC and Dell. Probably couldn't happen to two nicer guys. As Meg has said, that's going to go into some sort of swirling mass of cultural. What's her quote? She's been working on de-leveraged for how many years? Yeah, exactly. They're in a little sinking hole right now. I mean, there's a little bit, not a little bit, a big stuck-in-the-mud mentality. Yeah, and there's a lot of customers we're talking to who perhaps, maybe we're happy with their choice before or now actually giving HP a good choice because we look like the stable partner, right? So, I think that's happening. And I think the bottom line is, regardless of all that. So, they think you're more stable than Dell. Absolutely. EMC, because of their transaction. Right, but you know what? Eventually, all that'll sort. Sure, we'll have whatever it's success it has. Our job is actually to innovate for the customer and actually provide them capabilities that they don't have to go look anywhere else. So, if you look at, again, software-defined, we're there. Hyper-converged, we're there. Converged systems with free-par, we're there. Composable infrastructure, we're there. Well, it points to integrated. Integrated technology is not a pure play store. It's the question we are asking. This is where Dave kind of come in to kind of be nice with mentioning David Scott. We've been, we're overtly saying, can there be a pure play storage in this new market? Because that kind of limits a little bit, not a little bit, a lot of what you can do if the demand from the customer is, words like composable, agile, integrated, static, application, DevOps. A feel for more than just storage. Well, you know, you guys know I left the storage company with N in their name five years ago because I actually believed in the Converged Systems Store and I think I was actually a little ahead of my time because now what we're actually seeing over the course of the last year or so is that actually turning into reality. Is there always a place for pure play? Sure, not at the market cap potential and the sales velocity that we're seeing today. It becomes a much smaller market segment. It becomes a much smaller market segment, right? But we believe, you know, if I think about the acquisitions that we've done at HP, three-part was a good acquisition. Store virtual was a really good acquisition. We didn't know it at the time. It was a really, really good acquisition. It leaves us positioned to actually provide the customer that opportunity and that choice. So I'm not, the bottom line I guess is, I can hope for all the bad things to happen in the industry to all my friends out there or I can innovate, bring the products together, make it easier for customers to consume them and then deliver that business value and that's the one thing that we can control at Hewlett Packard, is our ability to do that. And then let the rest of the chips fall where they may. Well, that's interesting critical infrastructure and add-on. That's right. I mean, that's the end of the day. That's right. You're in the engine room, you want to be working on the main stuff. We want our stuff to continue to have the success that we're showing the market today. So what's around the corner, Bill? Share some final comments here. What's around the corner? Where do you see the market spinning in terms of customer next step? Where is this accelerating to? Where will the puck be in a few years that you guys see? So I think obviously, we can't turn our backs on cloud. I think cloud will continue to be important although frankly, enterprise customers in cloud still seems to be an oxymoron, right? Everybody's playing. No one's actually, as I know, is actually doing. I think composable is interesting. However, I think server and storage can be doing more things together as flash goes into the server. And storage is a persistence layer. I think you're going to see some interesting things happening there. I'm assuming we're going to talk about the machine at some point here this week, right? You're going to see more of that happening where the server can actually, you know, hint to the storage about next pages they're going to bring in together. And we're the number one server vendor so that sort of work is going to proceed. And I think that frankly, I think this whole computing model where server storage and networking is actually taking hold, I think that'll be the other thing that'll happen. So it's composable infrastructure, it's cloud and it's- And converge is mainstream. Yeah, I know. I don't think we're, you know- It's still early. But I think customers now don't have to be educated anymore. And I think that's where we were in the last couple of years was having to educate customers what converge systems was. And it's easier than polymorphic simplicity, clearly. But we don't have to educate customers anymore. Now it's sort of like, do you have one of these? Do you have one of those? Do you have one of these? Well, boardroom conversation is app development and I want to have new experiences and under the hood is- That's right. The engine room, right? It has to power it all. That's right. Well, I like to think I don't work in the engine room, John. But, you know, yes, I guess- No, well, the guys- You throw some Twinkies down to me in a couple of bowls. The guys powering the apps will be server storage. That's exactly right. It is, you know. That's exactly right. A nice engine to power that. I think the beauty is all it's going to come down to software. Engine room's middle of me sounds a little bit more- Yeah, it's okay. I'll probably have to answer that. Back to the engineers. Back at home. We'll say more so like the Ferrari. Engine of the Ferrari. Engine under the Ferrari. The BDM's got it. Less latency. More power. More power. Lithium crystals are- Bill, thanks for coming on theCUBE. Really appreciate it. This is theCUBE. We're here, Bill. We'll be right back with more after this short break. Come to the engine room here.