 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. We are here live in Las Vegas in Mandalay Bay in the hang space at VMworld 2016. This is theCUBE's SiliconANGLE's flagship program. We go out to the events and extract the signal from noise. I'm John Furrier, my host. John Troy with Tech Reckoning, our next guest is Vishpal Shan, who's the Senior Director of Product Management at HP Storage, HPE Storage, HP Enterprise. Welcome back to theCUBE. Good to see you. Hey John, good to see you. You guys, obviously a big partner with VMware in the ecosystem. Give us the update, man. All flash all the time. It's a flash crazy world now. Yeah, if you want to talk about flash, you know, so to your earlier comment about VMware partnerships, we work with the VMware community across many different areas, right? Flash storage being one of them, a key one just because many of these virtualized environments today depend so heavily on the storage. And flash makes it a very, very attractive option for people running virtualized environments. So talk about where it's all fitting in now with VMware for you guys. Obviously, you know, the three-part success story, Dave Vellante always raves about, oh, it's the best that goes to HP's ever done. Do the gift that keeps on giving, as he always says. But now with the all flash side of it, how's it impacting the data storage, data protection, all the integrated stuff that the customers are looking for. Is it changing the game a bit, or what's that? So I think, you know, if I may, there's the core of an all flash offering, right? And if you break down the core, you can say it's about performance, it's about affordability, right? And clearly when all flash started, performance was the key, then there was the affordability wave. And then there's even now what you would call data services waves, right? Where the ability to do snapshot, so quality of service. So I would mark those as the core. Then you could ask yourself. Those are table stakes. Sorry? Those are table stakes. There you go, thank you, table stakes, yeah. And then, so now the question is, if we look outside the core, because the core is pretty much understood today, right? There's still lots of things outside of the core. So for example, how do you protect a flash array, right? How do you do data protection, integrated with flash, because the considerations are different. Your performance is different, your application characteristics are different. So what do you do with data protection? That's one aspect. The other aspect is your infrastructure, right? Your host connectivity. You know, your bottleneck used to be storage. You'd eliminate that bottleneck. Where's the bottleneck now? Is it on your host pipes? And then the third thing I'd say sort of outside, the core would be, you know, there are newer environments coming up. Containerized environments are an interesting place where you may develop on one environment and choose to deploy on another in these cloud native apps. Again, how does a flash array operate in those kinds of environments? So outside of the main core, there's lots of very interesting areas to look at. Talk about HV Enterprise and specifically how about the flash, the data protection and the host side connectivity. Not so much the storage. Talk about the difference of those areas and how they all work together. Yeah, so let's look at the data protection first, right? And so what are the attributes of data protection that matter in a flash environment? First of all, how often are you taking your data protection snaps? For example, are you using in snapshots? Do you go direct to a backup device? What is the latency impact in taking the backup? What happens to your backup windows? How do you restore quickly? If you're snapping every hour and the hour, do you go back with a full backup, apply incrementals? Can you do synthetic fools? So lots of different elements here. And I think the point of view is you could take backup from a point of view where I've got to back up my entire environment. I have EMC arrays, I have IBM arrays, I have HP arrays, I have a whole environment here to back up, right? Or you could say, hey, in my flash environment, how do I ensure it's optimized? Just like what Veeam did with virtualized backups, right? They took a very specific approach. I think the same thing can be said with data protection and flash. Do you see, so Pete, that's the story for primary storage. How do you, does that story change then as you're backing up to another flash device? Are you saying that in the field? So that's interesting you say that because you have different choice points now, right? So I could have two primary arrays replicating each other. Then I could be backing up the secondary array to a deduplicating device, that's one option. The other option is I could be having my primary array backing up to a deduplication device and replicating at the deduplication level. Or the device level here. Or I could be replicating at the host level. So I think there are different choice points. Question is, how do you choose one versus the other? And they're trade-offs, right? They're sort of pros and cons. And you want to be able to offer the customers that choice as well as the guidance as to when you would do one versus the other. I love the way you were talking about generations. We've gotten to this core system now of this generation of solid state. But there's all these other technologies coming down the pipe. We talk a lot about NVMe and connectivity and we talk a lot about 3DX point and that's going to change everything. Where do those fit into this framework that you've been talking about? So you go back now into the core and look at performance, right? Because there's got to be a performance.next. That's our industry. It never stays the same, right? Things always move. And so the key to looking through those technologies that you asked about, John, is to look at sort of the end-to-end path of an IO. And it starts from an application. It traverses some kind of fabric. It gets to what I would call a controller fabric on the storage side. And then from that controller fabric, whether data is processed, de-duped, compressed, for example, it gets written to back-end, right? And so you have to look at that end-to-end path. So some of the technologies that we've been talking about talks about different points here. So NVMe as a back-end connectivity for back-end media to the controllers, that significantly cuts the latency down. Now, but if you look at the latency envelope today, the lion's share of the latency is not with the SAS protocol back-end. It's with the media, right? And so if you did NVMe, you want to pair that up with storage class memory to get the benefit of that latency. And then you want to ensure as well that you are poking, say, NVMe over fabric to your host so that the protocol delays there go away. And so again, here you can see how NVMe impacts choice of media, choice of host connectivity. So you get that end-to-end IO optimization. Talk about what's next for flash performance, specifically across the host fabric, controller fabric, and the media back-end fabric. Yeah, so I think you have to then figure out now, and as in all emerging technologies, there's probably going to be different choice points, right? So if we look at the host to front-end storage port connectivity, that traditionally has been fiber-channel. We are seeing a rise of ice-cozy and ethernet. So the question is, what does that do with when 25 gigi ethernet comes to play, right? Do we see a shift there, a tip there? Maybe, I don't know. I think, again, you want to be able to offer choice points. And if you can reduce that whole latency using ethernet technologies, I think there's going to be a segment of the market that's going to be very attracted to it. We've been diving down deep into the technology stack. I'm curious if you're seeing the buying center shift as we get to more integrated virtualization teams, cloud teams, do you have to talk about these technologies now to them, and do they have to understand how to buy storage? So yeah, so that's a very interesting point because there is a segment of the market that says, hey, I am looking at a VM level or an application level, right? And I don't want to associate all the different component metrics. So I think that's a growing trend. And hyperconvergence, for example, is a perfect example of that where people want to look at it at the VM level or even at the application level. And as we get more and more entrenched into lines of businesses wanting to develop key competitive capabilities, we need to be able to do what exactly what you just said. What's the HPE story? Now that your HPE storage is an important component of what you all are doing. I mean, in relation to what John was asking, what's the future looking like that you guys are talking about in terms of your storage platform? So the opportunity for us is to bring the collection of different technologies to bear on our customers. And I view it as two things. So job one is for us to be the best storage vendor out there in the world. If I took that storage myopic view of things, right? But we're not a small company, we're a large company. And so that's a job too that says, how do we, the storage, and the server, and the networking and the compute play together, right? And so we got to bring the one plus one plus one equals five story. And that to me is the opportunity HPE can bring, right? Whether it's things like composable infrastructure where you can say, look, I have one set of infrastructure for our mission and critical applications. One set for my cloud native applications. Why should I have two infrastructures for that? I should have one infrastructure that allows me to compose the elements as I see fit for those environments. Some of them have different attributes. I shouldn't have to have different sets of infrastructure to do both. I think to me that's a great opportunity we can bring to our customers. Talk about HPE Enterprise now and storage. Give us the update what's going on in the business. Obviously the VMware ecosystem been strategic. You guys, again, like you mentioned, been there for a very long time, been a big, big, big partner of VMware. But how's business in general at HPE Enterprise? Storage, what's the update? What's the shiny new toy? What's the, where's the meat and where's what's going on? Yeah, so from an HPE storage perspective, clearly all flash is one of the rock stars there. We're doing great with all flash, good traction. We're seeing a lot of interest around software-defined storage and hyperconvergence. And it's interesting on the software-defined side, we've taken the same approach as we is, well, taken on the primary side, because we offer now what we call a common data fabric where you can deploy software either in running on a ProLiant server or a Blade server. You can deploy that same software as an appliance, if that's how you want to consume it. You can deploy it as part of a hyperconverged packet. We even offer it as part of our Healy on OpenStack cloud distribution, private cloud distribution. So again, bringing one technology, one offering that can span multiple shapes and form factors help make it simple for the customer. Otherwise, they're going to do or deploy 13 different things. So, final question, Vish. As a veteran of the tech business industry, obviously, storage is your focus. Here at VMworld, what are you taking back with you home as a key walk-away item from VMworld? Share with the folks what you're learning, what's the vibe, what's what are you going to take home with you as a walk-away proof place? VMworld's always been a great show, right? It's probably the one place where, you know, it's got such a rich ecosystem of vendors, such a rich ecosystems of offering, both complimentary and competitive. So, you know, we have this term we call frenemy, right? You are a friend in some places, an enemy in others, which is great because it just gives you places to collaborate and give new capabilities to your customers. The vibe's great at VMworld, a very rich ecosystem. They're doing a lot of the great technology innovations in cloud and software-defined. We partner in main spaces, we compete in some, but hey, that's just the way the cookie crumbles. And customers want choice. Vish, thanks so much for sharing your insight on theCUBE, great to see you. We'll see you at HP Discover coming up in London in December. Yes. Right, I think it's December or is it? That is great. November 28th. Remember, okay, and yeah, right on there. Yeah, so big events, the European version of HPE, Discover, which we just had an amazing set of interviews. theCUBE was there, go to silkenangle website.com or youtube.com for a silken angle, check out the HP Enterprise Discover videos, tons of storage videos with all the big dogs on there. Thanks for spending the time here at VMworld. Thank you. We are live at the Mandalay Bay in the hang space at VMworld 2016. I'm John Furrier with John Troyer with Tech Recording. We'll be right back. You're watching theCUBE.