 Welcome back everyone, we are live in Silicon Valley in Santa Clara Convention Center. This is the open networking summit 2014, hashtag ONS 2014. This is theCUBE, our flagship program where we go out to the events and extract a signal from the noise. I'm John Furrier, the founder of SiliconANG, I'm joined by my co-host for this segment. Stu Miniman of wikibon.org, analyst, chief analyst over there at networking. Our next guest is Rajiv Nagar, group program manager at Microsoft, covering the data center networking platform division. Welcome to theCUBE. Thank you for having me over here. I love, we always love to talk about platform wars in the media, it's like a platform wars. Microsoft has had a platforms and tools group for many, many generations. You guys have your big cloud division, you have a lot of things to think about across multiple device, Windows servers, a variety of generations of tech. Now with cloud and applications proliferating at the edge, where it's Internet of Things, huge shift going on. Really the cloud and the changes at the top of the stack are forcing the under the hood conversation to be revisited. So I want to get your take on one, that mega trend around the future of networking and has that vector into the customer's challenges at the data center level. Because it's software at the heart of the value proposition in a pre-existing market trying to manage a future of diverse clients, diverse applications, and at the end of the day deliver scale and results. That's a great question actually. So let me talk to you a little bit about the way we approach our development. As you are aware, Microsoft delivers services at very large scale to a very diverse set of customers. Our services encompass the consumer market, which is for example, as you know, we have Search with Bing, we have Office 365, which goes to enterprises, and we have Xbox again to consumers. These services are delivered to hundreds of millions of customers a day. A lot of real time challenges and some interesting challenges to deal with. In addition to that, as you know, we are cloud first. We operate some of the largest data centers on the planet with Windows Azure. In fact, you could think about maybe three or four other companies like ours that operate data centers at the scale at which we do. Just to give you a sense of scale over here, we are talking about tens of thousands of physical nodes, and I'm probably underestimating that. We are talking about over 250,000 customers, over 1,000 new tenants brought in to onboard it every day, and tens of thousands of networking changes on a daily basis. Where am I going with all of this? When you're operating delivering services at such large scale and operating these data centers, you cannot do this in a rigid, inflexible, manual manner as you used to do in the past. For us, we live, eat, and breathe the notion of a software-defined data center, which encompasses software-defined compute storage and networking. We take our learnings that we get from delivering these services at such large scale, and then we translate those learnings into products that we bring to customers that they deploy on-premise. Our approach is cloud first, learn from our experiences in the cloud, bring them to our customers, and we think that's the future. I want to ask you before I know Stu wants to jump in, let's get technical real quick, but I want to ask you to share with the folks out there why this show is so important. We've talked about the future of networking. Given what you said, the scale you're delivering, I mean, Xbox is impressive by itself, Azure and Bing, huge infrastructure Microsoft has, but given all that, for the folks out there watching that aren't at the show, why is this show important? What's going on with networking? What is the key disrupting opportunity or to be disrupted, potentially on your end, or to be the offensive, to be a disruptor? Why is this show so important? Why is this future of networking conversation relevant? That's a great question. So we are of the opinion that if we want to deliver great disruptions, bring them to the market, you cannot do that in isolation. You have to collaborate. Think about software-defined networking per se. There's a lot of interest around the customer's desire to avail of the benefits that come with software-defined networking, however, the market is still in its infancy. It's in order to help our customers obtain benefits that they will get, that we realize from deploying software-defined data centers. It is almost incumbent upon the vendor community to come together to collaborate to foster innovation and then deliver that innovation to our customers together, even as we focus on differentiated services in the future. But we can foster a virtuous cycle of innovation by coming together, and that's what organizations like ONF and that's what shows like ONS do. So to us, it is pretty important. So Rajiv, there's no doubt that Microsoft is one of the largest software companies in the world. It's great to see you here supporting the community. I know you were a keynote speaker at the Open Daylight Summit. To be honest, when we look at some of the bigger companies here, some people would kind of scratch their head to get feedback from you and say, this is all open source. Can you give us a little bit of insight as to how does Microsoft look at the open source community? I know you're involved in Open Daylight, so can you give us a little bit of insight? What is Microsoft contributing back to the community? What code do you do? What resources? And how does open source and Microsoft go together? That's a great question. Again, it all boils down to collaboration for us. So I am unsure that you're aware that Microsoft actually participates in a variety of open source initiatives. We deliver complete solutions to our customers. We manage some of the largest data centers in Windows Azure, and we deliver a great end-to-end solution with Windows Server and System Center, and then applications that run on top of that. But we participate actively in Open Daylight, which you mentioned. We participate, we just announced recently our participation in Open Compute. We've been participating actively in ONF, almost since inception. In fact, a colleague of mine, Albert, is on the board for ONF over here. We are participating actively in the ONF, Northbound API Development, where we are talking about applications, leveraging a common substrate or a common platform. We are active contributors to the Linux community. So Microsoft participates in a variety of open source forums in order to promote this level of collaboration. So this is not unusual for us. It's simply what we do on a daily basis. Yeah. So, you know, first of all, we were at the Open Compute Summit, and it was great to see Microsoft taking those server designs and donating them to the environment. That was real IP, by the way, too. That wasn't just like throw away, it was real IP. So Microsoft has a little bit of hardware, but you're primarily software. So you're involved in Open Daylight. The question I have for you is, there's a lot of concern in the community, is will there be really an open Northbound API so that we will be able to create applications that can go on a variety of clouds or hardware platforms. I would think it's in Microsoft's interest that the applications that Microsoft creates can go anywhere, but you want customers to have flexibility and get lower costs. What's your take on where that development is and what needs to happen going forward? Think about it. Platforms are in our DNA. If you think about Microsoft, the first thing that comes to mind is we deliver a platform and then we deliver the tooling for developers to be able to create or deliver great innovation on top of the platforms we deliver. So this is something we understand. I think there's a very famous clip of Steve Bomer running around saying developers, developers, developers. Speaking specifically about the networking domain, as we all look at the innovation that is coming about with SDM, we believe that the key component of that innovation is the set of applications that are being built and that will be built in the future in order to leverage the or harness better the power of the network. That could be integration with key mission critical applications today. For example, I do not know if you're aware, but our link team had demoed about a year ago a phenomenal prototype where we showed quality being affected in real time when our link communicated with the network with the underlay in this case and basically spoke about the importance of the call, the problems they were having, and then the network could respond to the demands of the application. So whether it's mission critical applications like unified communications, whether it's next generation applications that are built specifically along to deliver on better security, monitoring, digestibility, so on and so forth, these are the key components that will fully unlock the power of SDM. And so when we look at ourselves and we look at the fact that this is a key DNA to enable such applications, that's what you'll see us focus on in Open Daylight and other communities like this. So that's exactly what we do. Okay, I'm curious at your thoughts, what white space, what new opportunities do we have when we create these new open source communities in these platforms? I think it actually benefits everybody. It benefits our customers collectively and it benefits the vendor community. Basically what we are doing is we are saying together we can deliver greater flexibility, greater agility, greater mobility to the mission critical workloads or applications that our collective customers wish to deploy. If a customer wants to onboard their new application, it doesn't take you four weeks because the network has to respond. It can now be done instantly. If the customer wishes a certain level of quality associated with the application, that can be provisioned dynamically in a heartbeat. If a customer desires greater visibility into what is happening end to end from the host through the storage across the network, that can be enabled. These benefits can be delivered regardless of where the customer chooses to deploy their applications across any cloud and this can become a reality if all of us collaborate. I look at that ability for the vendor community to come across and deliver that level of innovation so that customers benefit as the new frontiers. Rajiv, I want to ask you because you've given your background, I'll say technology, you've done a lot of great technology innovations. I'll say you're leading a big group at Microsoft, giving the history of Microsoft the size of their platform. I want to ask you about the state of change. The mega trend right now is massive cloud driving this application. The economics, the innovations are really the lever of what's going on here. So you have old incumbents like Microsoft out there, you have new incumbents, VC-backed, Series D financing, they may not make it, and then a slew of new startups. So it's kind of the landscape. The goal of the old incumbents is to be disruptor, not be disrupted. And the new incumbents is actually find a position, not to die, make it given the change. And then the new startups is to find some white space to innovate on. So given that landscape, I want to ask you the following question. At Open Compute, we saw this tinkerer mindset, almost like the homebrew computer club kind of mindset playing with hardware. You see Raspberry Pi, you see Bitcoin mining. You're seeing geeky coolness going on around building hardware with new software. With cloud, you're seeing a new developer mindset. So for the new startups out there, for the mindset of the eyes who want to find a white space, what do you see as opportunities at storage? Where in the convergence of the platforms do you see an opportunity to really innovate? For someone to say, I don't want to get caught in the storm of big guys trying to scale the market. Is there a spot for startups? Where do you see the innovation? Is it a new software paradigm? Is it a flash memory with addressability? Is it working with new platforms? What's your take on that? It's kind of a loaded question, but I want to get more of your entrepreneurial take and technical perspective. I think it's all of the above, but let me take a step back first. The moniker old might apply to me, it doesn't apply to Microsoft, okay? So, yes, there are incumbents out there and the incumbents will continue to find a way to deliver great value to customers. The way I look at it, just directly looking at what you were asking, we are in the infancy of a revolution that's occurring in terms of defining the next generation data center. And that spans all pillars of computing. You referred to storage, you referred to networking. We are looking at a amount of scale that dictates essentially a level of scale that requires us to examine the economics associated with that scale very carefully, whether that's cost of acquisition, whether that's cost of operation, whether that's the sort of services that we will monetize or deliver internally within our organizations in order to help the businesses do what they do best. So, when we look at the fact that we are in the infancy of this revolution, there's innovation happening in hardware, there's innovation happening in software, there's innovation happening in terms of new platforms being defined, new APIs being delivered, applications being constructed that look very different than applications that were constructed in the past. It seems like it's fertile ground for everybody. In fact, I would be hard-pressed not to find opportunity in the age that we are living in. So, I think there's plenty of white space for startups and there's certainly enough opportunity for the incumbents to go continue to deliver great value. And let's decouple two things. Let's decouple the technological innovation that's happening that has great value and then how that value will be monetized by either existing players or the new players in the market. And frankly, value monetization occurs when players find a way to deliver that value in the most expedient, most efficient way to their customers. The technological innovation is welcomed by all four of them. I think that's, I mean, music to our ears, we preset all day long. We believe it's fertile ground. Love the infancy and revolution. Open Source is a big part of that. I mean, we've lived in a generation where it's been plentiful code. Now you see Open Source go a whole other level. With things like Open Daylight, again, more innovation. No lock-in. I don't smell lock-in anywhere. There's no lock-in anywhere. So the question is, is lock-in the scale piece? Is scale the new lock-in? Is there a lock-in? Is it ephemeral? Is it, you know, virtual? I mean, some differentiation at some point, you can't be commod... You can't commoditize everything. At some point there has to be a switching cost barrier. Is scale the new lock-in value proposition? Or, I mean, the whole world can't be commoditized. Value creation has to come from somewhere. So let's kind of separate the notion of lock-in from the notion of value-delivered. Lock-in's a good word for profits. I mean, lock-in, I mean, the old way was proprietary, but that's kind of gone. You can't take that away right now. I have the risk of repeating myself. I think customers value solutions that are delivered, services and solutions that are delivered by whoever it is that's delivering them, and they will pay for those. Whether it is low latency, whether it is very large scale, whether it is ease of provisioning, whether it is the platform capability to be able to deliver on what the application means. All of these dimensions are dimensions that customers will continue to value and will continue to pay for, and that represents opportunity for us to go actually execute against that and how the mechanics of that as to how the code comes together and is actually put together and how services and code come together to deliver a solution to customers, but that's our challenge as the vendor community. So Rajiv, my last question, I actually got some of our audience reached in and they said, what's the update on NVGRA? Because that was some place where Microsoft moved forward in the networking space, obviously the hypervisor layer, Microsoft's the key player, can you give us an update as to where you see that fitting into the whole ecosystem? That's a great question. In fact, I would rather that the entire community forgot all about mechanisms like VXLAN, NVGRE, STD. Those are encapsulation mechanisms. Let's focus on the value creation. Within Microsoft, for example, and Hyper-V, we deliver a very rich overlay network, virtualized networking implementation. We have a distributed virtual switch that's inherently extensible. We have orchestration, management and control through a combination of system center and a Windows server. We have partners delivering on extensibility, including implementing open flow as extensions to our distributed switch. Where am I going with this? We have a great solution that we bring to market that's in market right now with Windows Server 2012 R2. That's an end-to-end solution with a built-in gateway and also a rich ecosystem that comprises Silicon partners, system vendors, ISVs. It goes through the entire gamut in terms of our partners who are contributing to the solution that customers can deploy today in the box. NVGRE is a on-wire encapsulation format. No more, no less. Certainly, I don't think worth discussing anything beyond it's a mechanism. All right, Rajiv, thanks for coming on theCUBE. Great moves at Microsoft. Satya Annatel, new CEO in charge, making moves. Big fan of his. He's a CUBE alumni as well. He was on us with us at the Stanford. He's a great executive, he's very cloud, he's very technical, exciting to see his leadership there. We're going to see some cloud. I love cloud first. It just plays off the whole mobile first thing, which is good, good marketing. But you guys have a big infrastructure. Thanks for sharing your personal perspectives on the revolution and the data center. Final word for you. Put a bumper sticker on the future of the innovation around the data center and for developers out there and for folks who are looking to jump in and really contribute to the community. What is your bumper sticker about this marketplace and what would you share with the audience? Embrace change, it's here. Okay, we are here inside the CUBE covering change, covering disruption, covering the innovation strategies from the top players in the industry. Tech athletes like Rajiv will be right back with our next guest after the short break. We're going to do the next guest. Chief scientist coming up, all the geek actions, the future of networking here at ONS 2014. This is the CUBE, we're right back.