 Live from the Mandalay Bay Convention Center in Las Vegas. It's theCUBE, covering VMworld 2016. Brought to you by VMware and its ecosystem sponsors. Now, here are your hosts, John Furrier and John Walls. Well, good morning and welcome back to Mandalay Bay here at the Convention Center. The theCUBE continues its exclusive coverage of VMworld. I'm John Walls along with John Furrier. Glad to have you with us here for day three of our coverage. Great lineup of guests we have for you and one of those has joined us here, Paul Durzan, who's the VP of Product Management at HPE. Paul, good to see you this morning. Good to see you and thank you for having me. You bet. Tell us first off, your feeling, your take on the show, the vibe you've had here for two, three days now. It's kind of soaked in. What's your take? No, it's a great show, pretty alive, so lots of good things happening. As you know, HPE and VMware are pretty close partners, so I think it's really good for both companies. You launched a hyper-converged product a little bit earlier in the year. Walk us through that a little bit. Tell us about what new you're bringing to the table with that and what was behind that. Exactly, thank you. So we focused on a few things. Number one, a lot of customers have asked for a hyper-converged experience on the DL380 because it's one of the industry's leading servers, right? A lot of people have it and they really wanted that hyper-converged experience on there, so we wanted to start with a really good hardware platform and start to build on it. And then the second thing we wanted to do is we wanted to make hyper-converged as simple as possible. So we focused on the user experience because the goal there was to allow people who didn't have a lot of skills to be able to do basic things. So people, the skilled workforce could do what they wanted to do and not the mundane tasks. So we made a nice GUI where wherever your mouse is you can click on, drill down, get the information you need. We did things like incorporate the ability to create a VM in five clicks. So now you can go into this hyper-converged experience, create your virtual environment. You can actually update your infrastructure. So we wanted to extend hyper-converged from being kind of that software-defined storage layer and bring the infrastructure into it. So with three clicks, you literally can go and update your infrastructure. So again, all designed to be easy. So we have been big fans of the convergent infrastructure from going back to eight, went before HPEHP. You guys pretty much coined that term. But really the trend now, as David Floyd is pointing out on our intro, was the trend around what VM we're showing here today is about everything being together in a hyper-converged. Because the network now has to be faster. Everything else is getting faster. So this is really what customers are moving to. So we get that, and gradually you guys have a really great product. We're all impressed with it. But I really want to get into the difference between what the hyper-converged does. It's a great product, everyone should look at it. But how does it fit into the new composable architecture that you guys have been talking about? Because I love that, messaging. Are they together, are they work together, they separate, can you break that down for us? Sure, absolutely. So right now, we view hyper-converged as being on the path to composable. When you look at composable, again, hopefully everyone knows this, but there's three key features. How do I create a fluid pool of resources? How do I go and put a layer of software-defined intelligence over that, and then have a unified API so you can control the full infrastructure? So when you look at hyper-converged now, what it does is it brings in that software-defined storage layer. And within our product, we've actually brought that together with one view. So you have the ability to go and provision your infrastructure, provision your storage. And over time, what I think you'll see is hyper-converged in composable infrastructure starting to merge. Because again, hyper-converged essentially has that foundation of creating composable fluid pools of resources and having the ability to go in and have the software-defined intelligence. So composable, the way I understood the different HPE Discover when we did the cube there. All the videos are on youtube.com. So I still look at the angle. Check out HPE Discover. We've got a zillion interviews out there around this one topic. I want to just break it down again because I want to make sure I understand it. Path to composable, does that mean that the hyper-converge is a requirement for composable? Or is it's a one of the building blocks? Great question. So we view hyper-converged as being one of the building blocks. And it will get you to where you need to go. Again, I see long-term those things really intersecting. And you can imagine them becoming one. But right now, hyper-converged is a great path to start to get onto that path of composable or a great first step. So composable, would I see a true statement if I said, no, kind of dumbing it down, the messaging? Composable is an invisible infrastructure to the developer. Is that a fair, accurate, and says very? I think that's not fair, yep. So basically, you don't have to worry about where everything's sitting. You're just going at the resource pools, storage, compute, network, and programming a DevOps-like philosophy. Yeah, and you get infrastructure as a code where you can have software now reach down and kind of hit these bare-metal servers and automatically deploy them. And you don't really need that intervention when that provisioning happens. So infrastructure code is the agility, the cloud native. Exactly. So where are your customers with this path? Because this is really important because I love that. I think that's going to be the end state for hyper-cloud and be the foundation. What are customers doing right now? Where are they? If I'm a customer of HP, I might have older boxes. I might have some reliance from the old days or some stuff, kind of servers and stuff. Do I rip and replace? Is there extendability? How do I move to this fast? Yeah, so one of the nice things is we have a bridge between today and composable, and that bridge is essentially HPE OneView. So OneView is that software-defined intelligence. And today, OneView works on DL servers. It works on BladeSystem. So what OneView will do is it will abstract the personality of that infrastructure and give you the ability to have your higher-level constructs, whether it's SaltStack or Puppet or Chef Ransom will come in and program that infrastructure. So that becomes that common element as you move from today to the future where it does the same thing now, except now you're going to have a more malleable, underlying set of hardware to give you that composability. Well, what's the biggest thing that you've heard from customers? You're on the product management side, so you've got to look at the roadmap. Yep, absolutely. So you've got to make, there's always those decisions that don't make it to the market, and everyone's fighting internally. We should have this or that. What's the features that you guys have that you see coming that customers could expect? Yep, and obviously I can't talk too much about roadmap, but I can tell you philosophically, we know that simplicity and agility are key, right? So this hyperconverge experience we built onto 380 was, again, designed to be really, really simple. I think we talked to a lot of people in the press about it, and people actually tried it and said, wow, this actually is simple, right? That's some of the reaction we got. So I think you're going to see that simplicity going forward from HP. How do we make it as easy as possible to manage? And then how do we take this world of agility and make sure that that infrastructure is as agile as possible? And it can talk to the software above it so you can start to have this automated, no ops, invisible infrastructure world. Yeah, define agility here a little bit because we hear about it a lot. I get, we get simplicity, right? You know, three clicks, five clicks, great. I could do that, John could do that. I can do it. Wow, maybe John could do it. Maybe not me. But agility within the workflow of different clients of different needs, different companies of different priorities. So how do you define agility? Yeah, so the way I look at it, especially in context of composable infrastructure is when you have these fluid pools of resources, what you want to do is you want to give the application the exact footprint it needs, and then have the ability to expand or contract that footprint as it needs it. So you want to be able to, from the infrastructure, create a footprint from it for the application as quickly as possible, ideally automated. So you do need some human intelligence to go and create that initial set of policy. But once that policy is created, you want an application to be able to come in, reach down into the infrastructure, get the exact footprint it needs. When it doesn't need it anymore, you just disperse it back into the broader pool for the next application to get it. So your time to value or your time to get that service up running becomes as fast as possible because you can move quickly and you can create quickly. So can you give me a real-life example of that then? What would be something if I, like you said, I'm trying to get up and running, I'm trying to take care of a task. Now I'm done with it, I want to move on, I don't need it anymore. So what would that be? So we obviously work very closely with HPEIT, and they have a great example of what they've done with our hyper-converged product and their development. So they're using a Docker environment but on VMs, and it used to take them 28 days to provision the environment for their developer to get what they need. Again, because in the hyper-converged, you essentially press a button and you get the VM it needs, they put the right image on those VMs and they serve it up to the developers because we have role-based access. So the developers now come in and click the VM it needs and they've gone from 28 days to provisioning environment to two hours, right? So it's this ability to go and get things up really, really fast and really, really simply that is what it's all about. Well, I'll talk about for a minute, share with the audience your relationship with VM. We're obviously, you're a VMware partner at a significant level. You guys provisioned VMs in five clicks, as you said. This is a key part of it. So share the relationship. You guys kind of work you guys do together, the status of it and what it looks like going forward. Yeah, so VMware's obviously a very critical partner to us and what we want to make sure is we want to make sure we make things as easy as possible either through what we're doing or through integrating into key partners like VMware. So for example, if you look at what we do with one view, if you're in vCenter, kind of in what I'll call the more traditional blade system, three-par environment, you can go into blade system, provision your servers within that system, provision your storage and your volumes within that system, and then provision your VMs within that system and have everything up and running. Again, HPIT went from multiple days to hours to do this because we've integrated together. So a key construct for us is we always want to be able to accommodate the VMware stack where our customers want it and we want to integrate as deeply as possible and I think it's fair to say, we're one of the more integrated partners they have and that's the philosophy we want to continue as we go forward. So I'm going to ask you the tough question and this isn't one that I think you guys have a good answer for now or getting a better answer than it was a couple years ago. Oh, the way he sets you up for that too. No, I mean, no, because everyone knows about HP's cloud situation. They had a couple of missed starts. Bill Hill's no longer with the company and now they're back to their roots of their clear strategy of basically what Pat's talking about, inter-clouding or you guys don't call cross-cloud but HP's always said from day one, managed cloud was a big part of their philosophy. So I got to give shout outs to the HP team. Managed cloud has been there from day one, they've never wavered. There's been some initiatives that didn't work out and same with VMware by the way, so they're kind of built victims but they say the clear data center path is there. So I got to ask you, when a customer says, hey, I'm an HP shop and I got some Dell and HP's and all this other stuff, I just want to go to the cloud. I'm going to go to Amazon. So maybe I'm going to move away from the HP because HP can't give me that cloud like agile experience. I'm sure this is where, this is the key, the composable. Talk about that nuance and that kind of statement and clarify HP's position vis-a-vis offering that kind of roadmap to the cloud. Yeah, so I mean, we would agree, right? We all know the world, the world today is a hybrid world and you have to be able to accommodate private infrastructure and public infrastructure and in fact, the ideal is really to move between them based on a set of policy around economics or latency or locality or whatever, right? So that's exactly the viewpoint we have. How do you go and set policy up so you can broker across clouds and put applications where they need to? But obviously if your on-prem isn't as efficient as it could be or not fully automated then it will have a hard time struggling or competing against public. So what we want to do is make sure we bring this agility to on-prem and composable infrastructure is the foundation of that. So your on-prem could move as fast as agile and get that public experience and then you use a higher set of policies to determine where do I go. So if someone says, oh, HP's late to the game which by the way is not true, you've been doing the composable kind of, I'm kind of teasing it up for you because I want you to explain, I hate the fud on both sides. You know, HP has no cloud but you really do have a path to the cloud. We absolutely, thank you. Yeah, sorry, we absolutely do and in fact, what's interesting about HPE is we have two paths because when you look at the enterprise, there really is this traditional and this cloud native. On the cloud native side, we have our staccato and our healing open stack. On the more traditional side, we have CSA and the software stack that goes along there and then we have a service catalog like Propel that can link across them. So we can actually give you both that cloud native and that traditional environment. And I think that's actually something really unique about HP because people either can't do the whole stack or they focus on one or the other. And there's no silver bullet too. I mean, one of the things I can say, go do all the theCUBE interviews, all the different shows, you know, there is an awesomeness around AWS but they're also trying to get into the enterprise which is essentially just Amazon and cloud and the enterprise, it's basically private cloud but Amazon private cloud. You guys have the data center footprint and again, access to the cloud and it's complicated, it's complex either way. It's very complex. It's trade off both functionality, performance, right? So you have, this is really the decision tree for the customer, right? It is, and you know, we do believe that the future is how do I go and move across these clouds and how do I broker across them? But that is a very complex thing to do and no one solved it fully but I think we're a good ways along. We've heard a lot of predictions this week, right? That's what shows are all about. You know, people get on the stage and make these predictions. In your world, the next 12, 18 months, so what do you see as being, you know, kind of like this is the trend that I feel is coming on. This is what we're going to be addressing and I think we're, our path to success. Yeah, so I think, you know, we talked about composable infrastructure and we talked about hyperconversion. I really see, you know, this, not only this melding of forces but how do I go and take this composability and make sure it can be automated or orchestrated. So it is really about this world of how do I get my policy around my application and drive that application onto my infrastructure with as minimal involvement as possible so I can be as fast as possible and I think we're going to actually start to see that happen in a much more robust manner. Well, Paul, thanks for being with us. We appreciate the insight and the time. Thank you, appreciate the time too. And we wish you continued success. Thank you, appreciate it. Thanks. theCUBE coverage here at VMworld continues from Mandalay Bay in just a bit.