 Okay, we're back. This is Dave Vellante of Wikibon.org and we're live at VMworld 2011. This is Silicon Angles continuous coverage. We're inside theCUBE. I have two great guests here with me. I'm pleased to share their knowledge with our audience, Prasad Rampali who's the senior vice president of EMC's solutions group and Mark Lesher who is a director of the virtualization practice in that group. Gentlemen, welcome. Thank you. Thanks for taking your time out to join us and VMworld 2011, the tech industry event, the enterprise IT economy, I've called it. So let's start, let me actually start with Mark if I may. Mark, what are you guys showing at the event and the solutions group? You got a bunch of demos going on at the booth. What's the action like? Well Ashley Dave, we're showing quite a lot over at the EMC booth. We have a lot around virtualizing mission critical applications. A lot of this is work that's done on the latest version of vSphere 5.0. So a lot of the performance, the million dollar IO work. Again the team there can talk about the testing that we did as well as some of the work around VAI integration. In addition to that, we're also showing some demonstrations of the vCloud director with workflow and the orchestrator. And what that allows us to do is allows us to automate some of the workflow provisioning tasks that a lot of our large customers are looking to try and do today. Great, Prasad I wonder if I could come to you and ask you to, we talked about this in the past in the queue, but I wonder if you could refresh our memory on the solutions group that you run. You've brought a new vision, a new energy, taking it to the next level if you will. What is the charter, the solutions group, and talk a little bit about what you're doing there? Yeah thanks, that's something I wanted to address. The charter of the group is to really deliver game changing solutions that differentiate a lot of our core storage technology and the emerging big data and cloud related technology capabilities. And central to that is really addressing endemic pain points in the hybrid cloud space, right? There are two or three of them that are well known, so I'll just touch on them briefly. The first as you would expect is this whole focus on end to end trust, on how you enable the movement of data as well as your core computing applications in a service provider cloud and ensure that you don't compromise on data integrity and the core security and compliance requirements. The second is quality of service. How do you ensure that as you move your applications and workloads to a public cloud or a service provider environment, you don't compromise on your core SLAs to your customers and you continue to hold to that. And lastly, the real focus on ensuring that from the time you have a need for a given application to be run in a public cloud, you can do that in minutes and seconds versus take a week or a month, which is onboarding, right? And so ensuring that these kind of pain points are met in a simple fashion is what we're focused on around game changing solutions. So as you know, the marketing always leads the deployment and we've heard a lot about cloud, private cloud, public cloud. I mean, it's a lot of cloud action going on. When we talk to the customers in the Wikibon community, they're most are pursuing a cloud strategy. They're starting to get on the curve, but they're not sold on the hybrid cloud yet. They're concerned about that. You mentioned several of these concerns. So there's a lot of hand-holding that has to be done and a lot of services that have to go into this. And can you talk a little bit about what you're doing in particularly that service's angle? Do you bring that to bear in the solutions group? Do you take advantage of that? Do you bring that to market? How does that all work? Yes, I'll take the lead and I'll let Mark also talk about the specific work we're doing in that space. Our approach to this is if we can't theorize on how you build a hybrid cloud, we have to actually go build it ourselves. And so what we've done is we've created service providers centric labs and we're working with well-known service providers as co-participants in the lab where we take their specific OSS, BSS context, the specific applications that they want to offer as a service and ensure that we can actually ring that out as a reference architecture end-to-end with our core storage environments. And so we are starting off with compute as a service, backup and recovery as a service, and VDI as a service as a three primary solutions that a lot of the customers are asking for. And the service providers are obviously trying to provide that post-haste as they move up the stack, right? So we are really ringing out a lot of detail behind these three stacks, and I want to have Mark tell you maybe how we are doing that. Yeah, so it was compute, backup and recovery in VDI. VDI as the lead vehicles, but we are also looking at healthcare vertical, enabling patient care imaging as a service. They're looking at oil and gas, and last but not the least, and Flirgil from NISI Tech will come and talk to you about how we are partnering with them and enabling analytics as a service, high-performance data analytics, which is in many ways is kind of the last frontier that we have to enable as a service, given all the issues you have with data. Yeah, I'm interested in that case study, so. Mark, go ahead, fill us in on some of the details. Actually, so carry on discussion around the hybrid crowd. I think there's a lot of the technologies are there today and are coming, and we see that with some of the later two work and some of the data center interconnect technologies, coupled with some of the security products. But I think you're right in saying that today, I think you need a good service provider that's done a lot of work around process and management. So there's a lot of good work that's done out there, and as the automation comes in, as we're starting to see today, it's net new products like BXLAN for security coming out of the Mware, some of the products that we're seeing out of the secure cloud and TXE-type functionality. I think that will then allow us to automate a lot of those functionality that we currently rely on process to achieve. Talk about automation a little bit more. The driver for automation, I mean, obviously, simplicity, there's a cost factor. Is there a quality element as well? Absolutely, and I think once, good process allows you to ensure quality. In automation, good code and good application development, good processing capability allows you to ensure that you have quality of security. So the ability to not only guarantee security, but the ability to then look at compliance and audit that, I think is an important aspect that we don't have today as easily as we can and we'll have in the future. So how do you, or how will you measure that? Well, today, it's a very manual process, and I think what we're seeing now is we're seeing some of the new security products that are coming into play that will allow us to go out, look at environment, take the red measure against the set of policies and expectations, and then report back compliance figures on that. So I think there's a lot of interesting technologies that are going to come out there, and I think that will help bridge the gap for customers who are a little bit concerned about how to cloud as we move forward. Yeah, just to add to this, you know, the quality cannot be done as patchwork. It has to be ingrained in the design in the core architecture. And if you take a look at security, right, the biggest issue we have with asserting security is how do you establish the root of trust? It is well known that the operating system, the application layers cannot be the basis of the datum for the root of trust because they are eminently hackable. And frankly, the only place you can go to to establish that root, I believe, is in the hardware, right? This is where we are working with Intel, where from their core instruction set enabled at the microprocessor level and the chip set, we can do a measurement through this technology called TXT, which is trusted execution. And we are integrating our core RSA product environment by which we authenticate to the hardware and do the measurement and then say, hey, the virtual machine is bona fide, therefore the application can be trusted, and then we let the application run as part of the authentication process. And getting at this deep rooted integration with establishing the right principles of how you establish trust is an example of how we think we'll design in quality versus checking it after a post facto and saying, hey, we have quality or not. By the time it's, the deal's done. Now, you mentioned a number of different types of solutions, compute, backup recovery, VDI, healthcare, analytics. Are these all hybrid cloud, what you would call hybrid cloud solutions, or is it a mix? By definition, we are looking at implementing first in the private cloud environment, but each of these reference architectures, we want them to be multitenant enabled, and we want them to be exported to a hybrid cloud environment, i.e. work on the wide area network, enable live migration and things like that, bursting and so on, that we want to demonstrate in the hybrid cloud. And so the progression is you're starting with the private cloud, prove it in the private cloud, and I presume you've done this for all of these solutions that we're talking about. All three of them, yeah, today we've done that in the private cloud space. Yeah, it's a compute, backup, recovery, and VDI. And the NICI story as well, right? The NICI story is still unfolding, so I'll let Virgil talk about it, but we are really looking at enabling analytics as a service, and the first part of that is really ensuring that we can stand up a analytics implementation in a service provider environment, which we have to do at this point. We are working through that. Okay, so we're clearly, where are we in the private cloud deployment today? Are we there, it's hardened? You can prove that, customers can go and kick it and try it and deploy it, is that right? We believe so, yeah. Yeah, for those three. For those three, yeah. And I'll let Mark speak to it. We actually have a very good demonstration of VDI and the benchmarks in our booth here, right? Where we can show that we can essentially look at a network boot storm that can be contained in, I think it's 30 seconds or less across 5,000 desktops. Okay. And so we've proven that stuff, right? At a level where we think it is scale. Now we are taking it to the hybrid cloud space and saying, can we enable that same level of performance and scalability where we leverage cash, where we leverage link clones and various technology elements in the private cloud? And can we demonstrate that as a service in a multi-tenant environment in the public cloud? So. Great. Did you want to add anything to that, Mark, or? I think when you go into the hybrid cloud, I think a lot of those pieces are there as well today. We're seeing that with a lot of the implementations that some of our partners are going into service provided space. So I think it's even moved forward comfortably with some of the right partners. But I think as the technology matures and something that comes in, we'll have a much broader adoption next year. So I mean, it's, you guys, sounds like you're tracking the market. I mean, you're deep now into the private, the hybrid, you have visibility on it, starting to really solve that problem. I mean, I think it's consistent with the data that we showed at the top. Anyway, we fully expect it. This is really going to take off. One of the things we've been looking at is understanding the data. That seems to be a really important thing for users, like the interdependencies across the application. I mean, is that something that you're focused on? Is that the right area for us to be focused on as observers? What are your thoughts on that as far as enabling the hybrid cloud specifically? Well, I mean, that's obviously, the interaction of data is very important, especially if you're bridging your private cloud to a public cloud. So obviously you need to progress in a very intelligent manner. You need to understand your applications, especially if you're going to go in an active act environment, versus more of an outsourcing out to a hybrid cloud model. But I think that's been the same rule we had years ago, which is you have to understand the interdependency of data. If you're going to back it up, if you're going to migrate it, if you're going to control it. Is it fair to say that the greater the interdependencies across the application portfolio, the greater the complexities, and those will be the last to go into the hybrid? Or is that not the case in your opinion? I think in many cases it will depend. I think the first instinct to answer is yes. I think people tend to look at the lower hanging fruit and what can I, the software development, what can I push out into the hybrid plainly? But I think there's a very compelling case for some of the critical applications that are very well containerized, where you know what the dependencies are and you're looking for either greater performance, greater high availability, or greater burst workload capability. I think those very particular applications can be focused on and taken to a high required, very successfully. And the business case will be there. All right, good. Well, Prasad and Mark, thank you very much for coming inside theCUBE, sharing with us your knowledge. That's what we like to do in theCUBE, and really appreciate you guys taking it. I know you're really busy. Tons of customers here, a lot of partners, so thank you very much. It was great to see you again, Prasad. Mark, thank you very much. Thank you.