 Okay, everyone, this is SiliconANGLE's theCUBE, our flagship program. We go out to the events, extract the signals from the noise. We're live in San Francisco at the Red Hat Summit, and we're talking about what's going on open source, cloud operating systems, open stack, you name it, we're on it, this is an amazing event. I'm John Furrier, the founder of SiliconANGLE. I'm joined by co-host Stu Miniman at wikibon.org. Our next guest is Sam Greenblatt, VP of architecture and technology and CTO of Dell. Welcome back to theCUBE, your CUBE alumni, previously with HP now with Dell. Welcome back. Thank you very much. We were talking prior to we went on about just the innovations around operating systems and systems, and now we're in a whole nother era of, essentially open source, really setting the foundation as a tier one resource for enabling huge innovations. And we're living in a massive inflection point, some call it a bubble on the consumer side, but really it's a recast of a whole new architecture. So I want to get your take on a couple things about what's going on at Red Hat. You guys announced a significant announcement with Cloud and OpenStack, and talk about your announcement and what that means for this continuum of computing, computing industry going from data center centric, now to a cloud. And what does your announcement mean? What our announcement meant today was, we are working very closely with Red Hat on a couple things. CloudStack's very important, but we don't see it as a cloud. We see it as an infrastructure as a service, so much that when you buy REL 7 or 6.5, you get, if you order it, the OpenStack can source code, not source code, but the binaries. What's important about it is, we've done OpenStack for a long time and everybody who does it looks at us and says, okay, we got it up and running, what do we do next? And that's always a classic question. And what's happened is applications have to keep up and make the cloud a cloud centric, basic platform. And so therefore we believe that OpenShift is a great pass layer. We also do Cloud Foundry, but OpenShift is where we're putting effort or wood behind the arrow with Red Hat. We're also working on containers. And I saw you had Docker on and Solomon's a character, but what we like about Docker is it gives true portability to anybody who has a Linux platform. And I'm sure he talked about it, he's gonna do Windows in the future. So that makes pure. No, I think that's a scoop here. We'll make that right down. Hopefully that's not a scoop. No, he didn't talk about that though, but he did go into detail on some of the container benefits. Well, if you noticed Red Hat bought a small company in Seattle that does container Windows, so with the partnership you might see that, so. Okay, so let's get into the news. So Dell and Red Hat announced a private cloud solutions reading here my notes. Notable bullet points are co-engineering with OpenShift and Linux containers. It's available for RHEL and OpenStack. So it's an enterprise pathway. So the cloud is a nice bridge and all the customers I talked to that have Red Hat, I always say, hey, I got Red Hat, I got some NetApp drives, I got some EMC drives, I got this company servicing it. It's not going anywhere, but I want some cloud, but I want to look under the hood. And so they go, great. So OpenStack has been that great hope. It's a bridge for them to continue out their mission. So take me through your perspective of that customer, CIO, major financial institution or vertical, where do they get going? Well, we're working with a couple major customers on this. You got to remember that the abstraction is really being done by Red Hat Linux. So they've is able to get the drivers, get the code, get it up and running. What OpenStack does is it uses the Linux more or less the HAL layer, the hardware abstraction layer. And to OpenStack, it doesn't know what's under it. Linux knows what's under it. So if you go up the stack, you got RHEL, then you have Rev, which is the virtualization, and then you have OpenStack. So what most customers are doing right now is we're working on a lot of things, especially with Red Hat and RHEL, of making it able to span, if you will, multiple servers, multiple server forms, whether it's blades, racks, or even systems on a chip. We don't really care as long as we can pick up the configuration and go across it. So when we talk to Michael Dell, and we're always enamored by Dell's engineering capabilities, and they do move fast. So talk about the co-engineering piece with Red Hat and OpenShift. What specifically are you guys co-engineering? What part of that is from Dell? Where is Red Hat? Where's the hand shaking going on? Well, let me go down the whole list, and then I can tell you, the first one we're doing is we're working in storage. I go up once a quarter to see the biggest banks in New York. They want to have an open source object store. They think Swift is great. They think Ceph is good. They don't feel that they can scale. So we're working with Gluster and with our Equalogic, which you know very well and compelling. And our engineers are working on how do we take those, those basically arrays and scale out and scale up technology and map it into an object store as well as doing block. So Cinder and Swift are important. We think Ceph's important. We think Gluster is going to help us piece a lot of the things that the OpenStack has done. A lot of the things that the OpenStack community really hasn't touched. The second thing we did is, as you know, we announced Dell Open Networking. We put Cumulus on top of our Dell Switch. No one else has done it. And there will be others in the future. And we are building out a complete solution with Open Daylight. We looked at other projects and like OpenContrail that Juniper put out there. But we are at the point that we believe that it has to be a software overlay. It can't be point to point. We've already proven it can't scale. So we're working very closely with their engineers in that framework. The third thing we're working on is basically management. We want to be, and I showed it on stage today, we want to be able to provision very quickly OpenStack's a bear to put in unless you have tools. And we want to provide those tools. We started out with Crowbar. That went, served this very well. We're now working with Puppet and ASM. We have a great user interface on it. And we think that'll be great. I can keep going. So on the tooling side, are you feeling good at the toolings there and is it open source or is it binary that you guys are contributing? Well, what we're contributing is the reference architectures, the templates that define how to do it. And we're also putting in a rules engine so that it can work in any environment. But the rules engine is Drouals, which is open source. And of course, the code itself is Puppet underneath that does the discovery. And we've taken on with Puppet Razor, which was originally an EMC project and we're using Razor to more or less be able to do configuration. So Sam, you've listed off a ton of open source projects that you guys are embracing. We're well familiar. We actually had Cumulus on when you guys made the announcement about the open networking. Can you talk a little bit about contribution from a coding standpoint? As you mentioned Open Daylight, the technical lead for Open Daylight said that code is point of the land. So how many engineers, how much resource does Dell put into the game of open source? We are right now in a bronze level for Open Daylight. We put a lot of skin in the game. Everything we're doing with Puppet goes back into the Puppet Forge, and I hate that name because it reminds me of Source Forge. Sorry, Larry. But Puppet Forge, and we donate all that back. So we want our devices to be as discoverable as anybody else's. We've written modules that handle equal logic. We've written modules that handle compelling. In the networking, we've already put back some code. We have Active Fabric Manager, which goes in, finds all the connections. In Compute, we're still putting modules in to optimize Nova or Compute. So we're pretty active. But what we want is, in the co-engineering, we want those contributions in the future to come from our partner that's going to be the enterprise version, which will be Red Hat. So I got a question from the Twitter sphere around your thoughts around EMC's Viper product. So what, I mean, different approach. I mean, can you comment on that or is that? No, I can comment on it. Software, Viper is an attempt by EMC to bring together six disparate product lines into a software-defined storage. We think it's a great approach. We think it'll help customers. But we believe that how you should do it is the way we're doing it with equal logic and compelling. We're merging the stack into what we call NextGen. That's obviously, there's hundreds of NextGen's. And we're going to keep alive both products but working on a single stack. Viper is an abstraction layer above it. And what we find with abstraction layers when you're dealing with storage, you got to really build a software hypervisor that's able to work with the hardware much more efficiently. So in terms of approaches, you think that's orthogonal to the direction of like an open stack or is it just more of their version of the truth, if you will? I don't think it's orthogonal. I think I wish EMC and that team very well and putting together a overlay to try to bring their storage together. But I think it's really, you're going to find that our solutions as they come out. And Ellen Atkinson who runs our storage will kill me if I make any announcements like I did for Puppet. What will happen is I think you'll see us taking a different approach but a more effective approach. What's your take on the clouds? When we talk to folks who are across spectrum of converged infrastructure, who are dabbling in with the past layer and ultimately have customers that have the requirement for analytics, mobile, bring your own device. We have application centric DevOps world. There's some systems challenges going on around the cloud. So I want to get your take on that because as CTO and someone who's been in the industry through many innovation cycles, booms and busts and it does move in waves. We are now in a major innovation cycle. There is some new changes going on that's going to create some enabling opportunities. So the enabling platform of cloud is presenting opportunities. So the question is I want to get your perspective and where is the stack baked and where is it not baked? Where are the white spaces? What are your thoughts on that? I'll give you one on big data. I have a meeting after now with another Hadoop vendor. Our biggest concern is Hadoop has its own file system. Anytime you have a separate file system for big data that spells somewhat some trouble. What we're trying to work with all the vendors is in one common file system. So that if you have block, object or even Hadoop they can all live together. Where I think this, when you talk converged we used to talk hard work. Then we started talking about software. A converged cloud is a cloud that has IaaS, infrastructure as a service, PAS, platform as a service, NBS, which is mobile backend as a service and SaaS. And it even goes above that, that then you have containers. It's coming together. Is it there today? I think you can start building it but it does take some elbow grease. Okay, so let's take your question about open stack. People can stand up open stack. What do they do with it is the question what are they doing with it? So I want to put the question to you because I think that's the next conversation we'll hear at open stack summit in Atlanta. Which I get to keynote again. And we'll be there with the queue. We want to have you on again. You're on the keynote circuit. I don't want to be on the keynote circuit. I'd like to be home with my wife to be honest with you. You come home, call us to queue, get the keynote. Stu, you're out. No, Sissy, what is that next level of innovation? I mean, because that's the real meat in the bone. The next, first of all, when you get up open stack everybody thinks it's going to be like VMware. And they're going to just put the applications into KVM and it's going to perform the same way. And it's the same with people coming from Hyper-V. And what we see is almost like when you converge. Convergence is not going to happen in one stack. It's going to happen across stacks. And you got to do it through. We want to make paths convergent across both stacks. We want paths to be, we want cloud foundry, which is a pivotal project and Paul Moritz is there. We want them an open shift to look very, very similar. So what we think the big challenge for us right now is not fracture the paths layer, like people are trying to fracture, tried to fracture Linux. Paul always reminds me of the Linux Wars 15 years ago. Paul was in it. I was a young baby at that time. I was in diapers. So we want to get the industry to put the customer first. And that's everything we do at Dell and stop the fracturing and the paths layer. We don't want to see sass fractured. And one of the things we're working with is to make sure that we're able to get the containers. We think the containers are going to be the next big technology. And one of the things that we do believe is with Ramesh, trying to wake him up, working with us, we're going to create a more comprehensive solution. So Sam, it's obvious you're excited and most of the people at this conference are pretty excited about containers. And I think back to the early days of server virtualization. And it took a few years for people to really kind of grok why this was going to be so transformational. I love to use the word grok. Yeah, so I was actually working for the company that acquired VMware when that happened. And to be honest, I don't think everybody understood what they were buying. Even though VMotion was there at the time and we knew that this was some amazing magical stuff out there. What is it about containers that get you so excited? Explain to our audience why this can be the next big thing. The difference between the early days when Diane sold it to that old company. What containers do for you is it gives you portability. Portability is what we've always promised. But having CENOS or an operating system, I'm not sure what Red Hat has called the next generation that goes into containers. What's important is that you can take that app and put it on any Linux-based system. Matter of fact, you know last week Docker announced with AWS. We think that's great. You know, you look at Docker and I'm not pushing them. They went from about 25 people in the open source community to over 500 today on Git. But we think containers are going to be the next form of virtualization. Will it replace it? Absolutely not. You're not going to see SAP virtualize into a container. But what you will see is applications that you want to put onto a PC, onto a mobile device into the next internet of things which I don't even want to touch right now because that'll be another five hour conversation. So. And we've got three minutes. We think it's great. So let's talk about the containers. So DevOps really is the culture now. And you talk about this container. What we see is this new breed of application developers are coming in. And whether they're young, new school, or old school that are coding faster, agile, whatever you want to call it, are looking at as a systems approach without being systems programmers. So I think what's interesting about containers that has our attention, I want to get your perspective on this, the DevOps ethos has been, I want to program software and have my infrastructure self-heal, self-form, be self-provision, basically take care of itself. So that's kind of the modern application developer. That's essentially moving up the stack. Do you agree? And is that kind of where? All right, absolutely. You know, if you remember years ago, everybody talked about Mapper when it was on Unisys, creating an application quickly and not worrying about the infrastructure. If you really look at, and I hate to use an old example, but what's happened is, it's really starting to happen with containers. I don't want the application developer worrying about anything but the business logic. And the data should present itself to the container and actually whatever comes down on should be able to be almost like a pipe into that platform to get the data. So I got to ask you the question. The old days always about abstracting away complexities and that's the beautiful thing about software and innovation is that the line of abstraction and what you abstract away from either coders and or other things becomes interesting because now if you don't have to deal with it, it's like what Intel processors were all about. You know, hey, there's a lot of complex stuff going on but no one ever sees it because it just works. Where is that line moving from an abstraction standpoint if you have an open source community, you have an OS, you have distributed computing. I mean, these are systems concepts as we were, as we're getting back to the systems conversation. So with all this going on, this global mainframe if you want to call that or whatever you want to call it, where is the abstraction line where you can harden a functionality? The abstraction layer should be at the application. And as you heard me open, the application is the center of the cloud. A classic example, you get a message on your cell phone. Do you know what tower sent it to you? Do you care? It works. Do you know what vendor actually provided it? That's what the application abstraction should do. Abstraction creates a better chance for Dell because the more abstraction you have, the more the architecture of the hardware has to change to make sure we can create. Okay, so in your professional opinion, if you believe that, if people can believe that, then that will create massive disruption from the old way, best and breed, all those ways you used to compete. How does that in your opinion change the landscape? Because it does disrupt, it's a disruptive paradigm. You're essentially horizontally scaling. Right. At the same time you're constructing and building more. So what's your take on that? We've been talking a long time about disruption. Disruptions only disrupting if you're the guy being disrupted. The other guy's the innovator. And what we see is people have to embrace a whole new way of doing applications. And when that happens, I think we're all going to benefit. So the disruptors are the innovators. Essentially people have to really figure out are you being disrupted or not, right? So Dell's an incumbent. If you want to look at quote, if you want to classify Dell and you're basically incumbent, you have computer supplier, computer systems vendor, you have software now, Michael Dell took a private, which is great. We're a big fan of that move. How do you guys maintain that culture of being a disruptor and not being disrupted? Because you have territory. You're on big market share. We have a big market share in the server industry. As you know, we're probably one of the leaders in web tech in the industry. That was a disruptive move about five years ago. Today, a lot of people are trying to get into web tech. What happens is how you do it is you start to look at the architecture. And I always get in trouble for saying this. The von Neumann architecture, which most computers are based on and our friends down the highway are involved, is going to change to more of a parallel compute architecture. When that happens, Dell will have the hardware and the software to meet those needs. So offers at the center of the universe. Final, I'll give you the final word in this segment. What, tell the folks out there, what is this event all about here at Red Hesse? At this point in time, obviously a 10-year anniversary, that aside, it's a nice fanfare, confetti coming from the sky. But from an industry, from a disruption and an innovation perspective, what is so important about this show this week? What's important about this show is we're taking the key open source components and Red Hat and Dell and everybody else is looking for someone to make it enterprise grade. Because if it's not enterprise grade, the banks, the major companies around the world aren't going to use it. So what's major about it is all the people can give the feedback to the leaders of Red Hat, the leaders of Dell, the leaders of OpenStack as to where the future should go. And there's no better place to do it than here. I'll just tell you one funny story. I was at the first Red Hat Summit and at that time, Paul had all these in Massachusetts. Yeah, he didn't want to travel. He didn't want to travel. Sorry about that, Paul. And today, it is snowing in Boston, so I'm much happier that it's here today. Sam, thanks for coming on theCUBE. Really appreciate it. Sam Greenblatt with VP, CTO of Dell, CUBE alumni, industry luminary tech athlete, as we say. Thanks for sharing your perspective. Congratulations on the Dell announcement of Red Hat on the co-engineering for OpenStack. Again, moving the needle OpenStack, becoming enterprise grade. That's what people want and that's what's happening. This is theCUBE. We're live in San Francisco at the Red Hat Summit. We'll be right back after this short break. Okay, thank you.