 Okay, we're back inside theCUBE here at EMC World 2012. We are doing a spotlight on cloud services, cloud applications, cloud deployments, hybrid cloud. I'm John Furrier, the founder of SiliconANGLE.com. I'm joined with my co-host. I'm Dave Vellante of Wikibon.org, and we're here with Toby Owen of Rackspace. Toby's involved in the hybrid cloud solutions part of the business. Welcome to theCUBE. Thank you. First time on. Yeah, it is. Quite a show. Have you typically frequented EMC worlds? No, this is our first time. Yeah, what do you think? It's impressive. It's a big show. Lots of good technology going on. Yeah, a lot of customers here. A lot of action going on, right? Absolutely. So tell us what's happening with Rackspace. The hybrid cloud, hybrid cloud's hot. We're going to share some data with you, and what are you guys seeing out there? Well, we initially launched a hybrid cloud offering called RackConnect about a year and a half ago. Right. And we've really seen kind of tremendous pickup from our customers over the last year and a half. Yeah, at the time of the announcement, hybrid cloud was sort of new, right? And the reaction was sort of tepid. People were sort of, I mean, we did some surveys at the time, single digits in terms of people said, oh yeah, hybrid cloud, that's my primary strategy. That's really changed in the past 12, 13 months, hasn't it? Yeah, it has. I remember attending some cloud shows a year, a year and a half ago, and hybrid cloud might have been on a bullet point on the slide, and now there's entire segments dedicated to it. So it's definitely- Why do you think that is? It's just the industry's got a better job of sort of proving it out or? Well, I think a couple of things. I think that as we sort of peeled back the marketing shine on cloud and customers are starting to really dive in, particularly enterprises, around how we're going to use this for production applications, they've realized that moving everything to the cloud all at once represents sometimes a pretty major project. And so hybrid is really a way to help customers start to use the cloud as a part of their application so they can start that migration a little bit quicker. And I think another reason it's important is that it allows existing kind of traditional IT, dedicated servers, dedicated storage to be paired up with cloud. And so it can help overcome some of the shortcomings that cloud that's still really in its infancy may have today. But I infer from what you just said that there's a direction that people want to get there eventually. So you see that as the end state is a predominantly public cloud strategy. Is that right? We do, I think cloud today, we're seeing more and more adoption but I think the bulk of all the workloads are still running kind of in traditional formats whether that's hosted or do it yourself in your own data center. And I think over time we're going to see that change where most applications are going to be running in the cloud but there's still going to be a need for dedicated infrastructure to kind of fill some of the gaps. I think hybrid is not a short term transition as a way to get to cloud. I think hybrid is going to be the reality for the next 10 or 15 years. So it's a business model you're saying? Absolutely. So how are customers dealing with to the consistency and policies and procedures and things like security? What's an incident? What should be reported? What can I audit? Is the industry stepping up to provide that level of visibility and consistency that is satisfying the enterprise? Is that a part of this sort of adoption? Kurt? Well I think the industry is responding to those needs. I think it takes time. Any new technology, any new platform takes time to develop those things and kind of the higher level requirements are definitely being driven by the enterprise today. We're involved in a project called OpenStack which is seeking to define kind of an open standard for cloud computing. Yeah we know it well, we've been following it. We're excited about it. We've put forth some caveats to our customers but still it's a lot of momentum behind the developer community. Yeah the networking side seems to be dynamic. I was talking to Nisiera. They're doing a lot of software defined networking as well. A lot of the network guys are coming into OpenStack so we were kind of commenting like OpenStack is getting traction down at the network layer kind of quietly. We're kind of talking about that. How does that affect your architecture? Because you want OpenStack to be open. You want the ecosystem to develop but you guys got a business to run service customers. That's right. So you can be open and grow in the ecosystem same time you got to deliver operations. So it's ops dev in this case, right? Dev ops dev. So how do you guys handle that challenge and talk about the balance of the networking component? Well I think networking is really key to building a cloud at scale. I think some of the earlier cloud platforms including our own really ran into challenges when we got to a certain size. And I think software defined networking is a big key to kind of unlock the next level of scale. And so implementing things like isolation within a multi-tenant environment are really going to help overcome, like you mentioned, some of the security problems or concerns around cloud today. From Rackspace's perspective, there's a lot of talk about convergence and you're bringing compute and storage and networking together. From your perspective, what does that all mean and we hear a lot about the good. What are some of the nuances and gotchas that you have to worry about with regard to that? Whether they're organizational, is it sort of new process models you have to put in place? How's that all sorting out for you? Well I think that's an important message to deliver that there are things that customers need to think about as they're planning kind of that journey into the cloud that it's not take my application and forklift it into the cloud and everything's going to work the same. And so I think one of our approaches and how Rackspace really are kind of key strategy and vision is to be one of the world's greatest service companies. And so our cloud strategy plays into that pretty well where we want not only great technical support but also great user experience. And so all of our design principles are geared around making cloud easy to use, intuitive, hybrid is another example there where we've really gotten down into the weeds and make that infrastructure easier to provision so that customers don't need to think as much about all the ins and outs of how the network works between two different environments. So what is a hybrid Metro? Tell us a little bit about that. Hybrid Metro is really a proof of concept that we've started in conjunction with Brocade and EMC as well as Masergy communications and it's based on a technology stack to do synchronous replication within a metro area so that workloads can be more easily migrated from a customer data center to a cloud provider. And we really feel like the next frontier for hybrid is moving outside of a single data center, a Rackspace data center where we've got dedicated in cloud sort of under the same roof and connecting up in an easy way and an integrated way to a customer environment. So hybrid Metro is the result of some testing of this use case, this technology stack and it's really a way for us to engage with customers and explore with them, is this a use case that meets some needs that you have today? Are there other use cases that are similar? Because we really feel like if we can do a good job enabling that connection to happen then we can start to build clouds that span distance and really start to see kind of a larger adoption of cloud. The objective is a zero data loss synchronous capability for disaster recovery? Yeah, I think there's a number of use cases. I think that zero RTO business continuity is certainly one of them. But it doesn't necessarily need to be that. The ability to easily move data through replication between sites makes things like live virtual machine migration much easier to do. The idea of I need to take a bunch of capacity and move it somewhere else because I've got some maintenance to do. Might be a great use case for that by shipping a large test of development environment out to a service provider so I can do this maintenance without impacting customers is another great use case. And then when you talk about doing it at a distance do you mean at asynchronous distance? Is that the next step? Or potentially, yeah. Right, I think we're really trying to get close to our customers on this to understand what are the needs they have today that are going to, or the pain they have around moving to the cloud and how will this solve that problem, right? So it's essentially putting the cloud into a remote replication environment. Utilizing the cloud, utilizing outsourced resources as opposed to building your own data centers. Exactly, so as enterprise IT shops are running out of space in their data centers they've got a couple of choices. One, sign another big 20 year lease and do construction and add capacity. Or two, might be to take those new workloads or additional workloads and move those to a service provider. So it's really, it's a different model for us in that we are now an extension of a customer's existing IT infrastructure. Chance to get out of the real estate business. That's right, that's right. So Rokey was just on talking about simplicity. What were they saying? Simplicity low cost, which is networking parlance for low cost, high performance solutions. But simplicity is a hard thing to crack in the cloud because it's not that simple. But as you move up the stack, the middle ground, middle layer is kind of complex. You guys doing that with multi-tenancies of the stuff. So what is your relationship with Rokey to talk about? Are they achieving that simplicity? I mean, it sounds good, but give us the scoop. I mean, what are they doing with you guys? Well, I think simplicity for the customer is what's important. I think under the covers, there's a lot of technology and development that needs to go into making a good customer experience. I think Rokey has been a great partner with us and really helped us kind of solve some key infrastructure areas. They've been involved in our IPv6 readiness. Some of the scale issues as we continue to grow are user base around really large networks helping us solve some of those problems. And I think in terms of hybrid, we've been actively engaged with Rokey as well in creating automation and integration. What does cloud-optimized networks mean? Because that's kind of a buzzword. It's kind of a fuzzy term. They kick that around. What does that really mean? Well, I think cloud-optimized network means that it's easy, it's provisioned in the same way that cloud is. Either through a user interface or through a programmatic interface and API. So that it can really be provisioned and configured on demand. I think if you contrast that with sort of the traditional model of let's stand everything up, spend a couple of weeks or a couple of months getting all the equipment online configured. It's a very different model, right? I want my cloud servers on demand at the push of a button. I want my network to respond the same way. Awesome, awesome. So DevOps, talk about Rackspace's vision around DevOps. I know you guys are very pro developer in this area. And it's not a surprise to go back to Lou Mormons conversations that we had in 2008 and Brett and those guys. So you guys are really ahead of the curve on this area. But now the EMC culture or the big guys are kind of getting a whiff of the DevOps. And I think next year will be the year of DevOps. That's our prediction, we believe that. To be true, what's your take on that now? Give a perspective on, from Rackspace's view and your experience at Rackspace, what's going on in the DevOps world and how's this crossing the chasm to the mainstream? Right, well I think, first off, we're following that DevOps model in our product development. So we've got developers and operations guys sitting together doing continuous testing. Allows us faster cycle times on development. And it really allows the production of better code in the end, right? Because you catch your problems earlier, you're able to address them kind of in line. I think from a cloud consumer point of view, it's, go back to what we talked about cloud optimized networks. It's really about having a programmatic interface to all of your infrastructure, right? And so you start to tear down those silos of system administrator, network administrator, developer because coding to an API. You got to do it all, you got to be versatile. In other words, developers aren't usually network guys and network guys aren't developers. So DevOps is really like saying, hey, there's no distinction, you're all one. Well, I think there's nuance in that. I don't think that the developers are going to, Not that easy. They're not going to come in and be a network architect, right? But the ability to start to provision infrastructure. Push buttons. Yeah. Well, coding and application looks a lot like coding to an API, right? So it allows some of that cross training and cross pollination to happen. I think it gives enterprises a lot more flexibility on how to deploy their internal technical resources. That's a good way to look at it. Dave, that's an interesting comment because what that does is that kind of gives us a frame because what DevOps really is this program to an API for a developer. Knowing that there's trust and then network guys giving an API, knowing there's trust because in the old model, it's like shadow IT exists because ops won't let the developers do their thing and vice versa. So DevOps right now is such a fun area to watch because depending on who you talk to is a different definition. But you guys are actually executing this in market with Rackspace, with OpenStack and then your product development. So DevOps, so we just did a survey in the Wikibon community. 40% of the respondents said DevOps term is new to me. You don't know what that is and then about 16% said we're just going to have to speed with Agile. Give me a break on DevOps. But at the same time, about 12% said we're doing DevOps and we're seeing hyper-productivity as a result. So it's new. It's a very exciting area for us. I think that's really the goal is similar to cloud. It's about agility and how quickly can you move through getting the next thing done in IT. We had a conversation with a customer recently that he actually took his systems administrators and sat them next to the developers to sort of force that model because they're doing more and more of their production applications in the cloud. And that's really helping them. A little empathy training? Yeah, that's right. So in one minute we have left. I want to just get your quick perspective. Really appreciate the time coming on. Rackspace, you guys doing some great work. We're big fans of Rackspace. We're a customer actually on one of our sites. Talk about security. Obviously it's a moving train right now. Securing the apps, securing the data. I just put some stuff out on Twitter just before you came on about the bring your own device to work is really kind of fueling the debate on how the interaction at the hybrid is. And so that's one factor. So talk about security from the edge. Bring your own device, securing the device which we don't think I'm not for that. Secure the app and the data and how does that affect the hybrid cloud? So the general tech consensus is secure the app and the data and you're good. And we've got virtualization to help there. How does that affect some of the cloud optimized networks? Yeah, well a couple of things there. I think as you move to a more distributed model which cloud really is, security needs to move down closer to the data. Right, you no longer have this perimeter network that you can define and then everything in the middle sort of that squishy middle with no security, right? So having kind of defense in depth and pushing that security closer to the data is definitely a good strategy. I think hybrid is a good way to improve security over potentially just a pure cloud environment because you can bring some of the traditional more mature security products to bear on a cloud environment, right? Having a dedicated firewall sit in front of your cloud servers or having DDoS mitigation or intrusion detection take part in securing your dedicated and cloud infrastructure together really helps kind of raise the possibility. Just the enablement of the brocade piece of that help you, they have good policy. I mean, how do they thread through that next layer? Because you know, that's essentially they got to have the policy. Do they have that there? Yeah, I mean, we support a couple of network platforms in our Rack Connect hybrid offering. And all of those devices are network devices, right? And so those devices aren't single purpose anymore. There's a lot of security built into them. And so that we're creating kind of that layered approach. We've got a firewall, maybe a load balancer applying security policies throughout those in an automated way through APIs through those network devices and brocades definitely helping us out with that. Okay, I think we're good. Dave, this is good for good segment. Yeah, so we really appreciate you coming on and sharing Rack Spaces, you know, thoughts and opinions and we're watching this space. You guys are in the heart of it, doing some really innovative things with OpenStack. So congratulations with that. DevOps, great success story. Thank you very much. Well thanks, I appreciate your time. We appreciate it. Rack Space, great success story in the DevOps movement. I think this is the year we're going to be, I'll predict right now, go out on a limb. This is the year of DevOps and it's going to be really about operational efficiency with developers really adding a lot to the table, tools and hybrid cloud. So we'll be right back with more spotlight on cloud applications, cloud services. We'll be right after this break.