 Hi everybody, we're back. This is Dave Vellante at wikibond.org and this is theCUBE, Silicon Angles, continuous production of VMworld 2013. This is the spotlight on virtualization, data protection, backup in a VMware environment, a critical topic. One that I said, I didn't feel like we heard enough this morning about on the keynote and this is sponsored segment by EMC's BRS division. It's been a great support of theCUBE over the years and I love these segments because we have an opportunity to really dig deep and go into the minds of executive Steven Manley is here, he's the CTO of BRS, he's a longtime CUBE alum and Deepak Mohan is the Senior Vice President of BRS focused on the product side. Gentlemen, welcome to theCUBE. Thank you. So Deepak, you're new to BRS? Yes, I am, yes, but not to the data protection world. No, we were talking off camera. You've started right out of college in data protection, right? Yeah, absolutely, yes. Well, once in data protection, you can't seem to get away from it, right? It's the hard problems that you like to solve. So let me start with you. So what attracted you to EMC and BRS? What brought you here? What did you see? Yeah, so I spent about 14 to 15 years at Symantec doing data protection and 10 years before that at another company called Cheyenne Software. So I've been doing this for over 20 years and throughout my career, I've been competing with EMC, whether it was the remote office deduplication with Avomar, cloud backup with Mozi or the deduplication appliances with data domain. And I was tired of losing. We can't beat them, join them? That's all the old saying. But now EMC has become the largest data protection provided in the world. And I see us to be in a very unique position to continue to deliver the future of data protection. I asked Pat Gelsinger when I first interviewed him, John Furrier and I, what's your first impressions about EMC? And we asked them to compare that to Intel, of course. And he said EMC is insanely focused on the customer. Of course, I know this from my history of just following EMC. But so what observations would you make, coming in as a newbie? So absolutely that's right that we are insanely focused on the customer. But my job is to deliver those solutions for the customers that will last them for years. And one of the things I've seen is the technology assets that exist in EMC, we can provide very, very superior data protection solutions. So just as I say that, I mean data protection is all about protecting primary storage and primary storage workloads. And that puts us in a very unique position to be the number one provider of primary storage and the number one provider of data protection. So there are tremendous opportunities in bringing these things together. So that's what I feel is like, and as you were talking earlier in terms of creating these information stores and being able to do things with this information, once again EMC has those assets. So I think the BRS and data protection, we can put a lot of these things together for our customers. So Stephen, I want to turn to you because you're relatively new to EMC as well. I mean, it's been a couple years now, right? At this point, yeah, compared to Deepak, I've been here forever. Right, you're a long timer. But I want to say, so as I said, you heard me say EMC is insanely focused on the customer and I can attest to that. At the same time, it's an opportunistic company. It's very execution driven. It sees an opportunity, it grabs it. The data domain is a great example. And so however, as a result, you perpetuate certain challenges in the customer base and you came in with a vision to address that, this sort of bespoke technology stovepipe challenge of data protection, very distributed. And you're bringing this vision of data protection as a service. So I want you to talk about that a little bit and talk about where we are and that journey that you laid out two years ago now at EMC World. Right, so the high level picture in this journey is pretty simple, right? It's that more and more the protection has to move away from what I'd say are traditional backup client architectures where I'd put my net backup or my TSM agent on a server and that would drive a backup into a very proprietary infrastructure and it was kind of a my way or the highway approach. And what we've seen is with the amount of data growing and the importance of that data, more and more the intelligence of the backup has to shift into the hypervisor, into the storage stack, into the application itself and really trying to leverage that intelligence. Now, two things can happen. Either one, that can proceed without the backup team and they can just be moved to the sidelines or two, the backup team embraces it and works on integrating with their application users, their hypervisor, their storage stack. And that's really been EMC's take on it is we've got to get out of that my way or the highway kind of mentality into that service approach mindset. So we talked about this a couple of years ago and last year we talked a lot about Oracle integration and how we've really been working with DBAs. Now this year really has been for me the year of VMware. So it's been about really leveraging the intelligence and plugging into VMware, whether it's the user interface we introduced with the new versions of Avomar and Networker that are straight there for the VM admin or the continued integration with VMware from a data flow perspective so that you get really efficient VM backups last really efficient restores. And then of course the new feature that we just added in terms of you can actually spin up a virtual machine, I do my backup of a VM, I could spin that up off of my data domain. So really enabling the backup team now to say, hey we can actually help you with disaster recovery with backup, the whole set of protection services. So if last year was the year of Oracle, for me this is the year of VMware. So I want to ask you about the sort of differences because we at Oracle Open World, we talk about it's like our man world for the backup standpoint. And VMware, the parlance is different, but your vision suggests that I can actually apply based on the service level requirement services to those different applications even though they're speaking different languages. So how do you see those two worlds coming together? So you know, I mean it's like all things, I was sitting with the customer just this morning and the analogy we used is really any IT environment is it's like a geological dig, right? So I might have some new green field opportunities that I'm setting up that maybe are even a link between on-premise VMs and say VCHS. And that's how I'm deploying some new apps for scale. I might still have some old databases, frankly that I've decided not to virtualize for a variety of reasons. And going back even further, I might even have some old infrastructure running SunOS for that matter, you know? And we view it as our responsibility to say, no matter where in that geological dig you are, we want to give you that ability to provide the centralized backup service. So I see them as playing nicely together where in the end we're all trying to service the app admin. And we can service the app admin directly like we do with Oracle through our man. We could service the app admin through the hypervisor layer or we could even service over time through the storage layer. It's really about giving people the choice so they can fit in with what's in their environment. So Deepak, I wonder if I could ask you a question from a product standpoint. So you've got a, EMC has a portfolio, it's a rich portfolio, it grew even richer with the data domain acquisition and then that whole business really exploded. I mean, everybody knows EMC's got about two thirds of that market space. So where do you see, what's the priority in terms of resources? Making those individual products better or making them all work together? How do you balance those and what is the priority? Yeah, so what we are doing really is we are moving from a more, what I call a singular architecture to more distributed architecture. So our storage architecture, for example, is becoming very, very distributed. Stephen, we call it the protection storage architecture. You got it. And this is what it pretty much does is it will allow us to store the data in their native formats. So customers now have applications whether it's Oracle or virtual machines or VMware or Hyper-V in the cloud, on-prem, filers and you can no longer scale going through a funnel to a single data protection application traditionally. So it's becoming a many to many type of architecture, right? So the control is in the hands of the application owners and in order to get your RPO and RTOs, you have to be able to send it back immediately. So that's where we are moving towards into this more distributed, many to many architecture. So what that will do is allow our customers to no longer have to really worry about the single threaded recovery's restores. At the same time, the backend, now more than the data store becomes an information store because in its native formats, you can stand up VMs immediately, you can recover or you can just enable a snapshot and be completely live with the machine. So as well as if you want to run analytics or e-discovery tools, you can do it directly on this backend architecture. So what we are really doing is, of course, we're keeping the current things going, but putting a focus on the future of data protection. Yeah, so okay, so you're spending money on this many to many architecture which adds parallelism. Yes. So you do many things at once because otherwise you're not going to be able to put all that data through the little teeny pipe, relatively speaking. On the second part of that, that native format and being able to access that, that's a business value discussion, isn't it? Can you talk about that a little bit more? Oh, it's huge, right? I mean, it's interesting because you go into an account and one of the things they'll tell you is, data domain is, I get 18 to one D-Doop or 12 to one or whatever level D-Doop they're getting. And one of the discussions we have after that is, but I still have other copies of my data in other places and it's great that you've D-Dooped my backups, huge win, but I still have 11 other copies of my data. So our goal towards moving towards really open formats is that's the next wave of D-Doop. If I can take the copy I have that's my protection copy and use it for disaster recovery, this instant access of a VM or I can use it for validation or test and development or deeper analytics, now I can start D-Dooping and really eliminating some of those other copies and get that next wave of space savings, next wave of cost savings, next wave of OPEC savings. So for me, moving to this open native format, it doesn't lock the customers in anymore and vendor lock-in to me is a cop-out from any vendor, but the second one is it really lets us elevate backup from where it is today, which is really an insurance policy, right? Whether it's health insurance, life insurance, however you view it, to something that can really have a more active role in your everyday business and start to drive revenue. Well, I think the great thing about Data Domain when it came out was it was solving a problem because backup was perceived as insurance and it was expensive and so it reduced the data, but I always felt like, okay, that's a one-time hit. Where's the business value? And I think that's the challenge that the industry has and so what you're describing is that today, the data is locked inside of a format, okay? So my question is, the lock-in point you made struck a question in my mind which is this whole software-defined thing. In a lot of respects, that software-defined is designed to keep things moving freely, agile, quasi-open, even though there's got to be a degree of lock-in for anybody to make some money, but what do you think about software-defined? Where does backup and data protection fit into that whole thing? So from my point of view, and this is really when Deepak talks about us investing in that next wave of data protection, a lot of what we're trying to do is drive what has been traditional backup and data protection value into the infrastructure itself, right? And that I think is really where you get into the software-defined is that instead of setting up a separate process for backup, instead of having separate mechanisms, if you think of it, I deploy a VM, but now I deploy a protected VM. I deploy a database. Now I deploy a protected database and I make that part of the infrastructure, I make that embedded. That basically gives me that flexibility now. Now I'm not trying to coordinate between this group over here that set up the database and this group over here that's protecting it. When they really come together, that's that flexibility that we're looking for. And that to me is when really, again, backup stops being such a painful task. Yeah, so you deploy that protected database or that protected application or the protected VM based on the requirement that the business sets, the SLA, the RPO, the RTO, or even- You got it. Even high, medium, low, whatever it is. It's okay. So where are we in terms of being able to deliver that level of functionality as part of an IT service catalog? So one of the really great things we had, again, in our latest announcement on the backup software side, is that you can have policy-defined protection layers. So for example, one of the big things that we've seen in a lot of accounts is you'll walk in and they'll say, hey, did you know we back up 1,800 VMs? And then the VM guy will raise his hand in the back of the room. Yeah, we have 3,000 VMs. And that's that, oh boy moment, shall we say. If guy were here, he'd use the different word. And what we're really trying to shift to and what you can do now with the new Avamar networker is, you know what, when you create that VM, it's automatically protected. And so in that range of the service catalog, it's already there. The next wave that we're also talking about, again, as VMware talks about VCloud Director and VCHS is, you know, the world's gotten more complicated. It's not just a database anymore or a file server. Your application is usually multiple databases with some glue codes and files. And with V apps and things like VCD, you can actually tie all of those together and we can help you not just back up each individual component, but treat them as a whole. So again, I'm setting a policy on my real application, not just the components. So Mohan, your former company has a big share of the backup software market, right? So you've got to play in that world. What are your thoughts about, you know, leveraging that capability? What are customers telling you in that regard? No, I think as we look into the future, as I mentioned, I think EMC is in a very unique position to deliver this story for tomorrow. And as we heard Stephen say, this next generation where we are trying to protect primary storage applications, EMC has those assets. So we have some very interesting roadmaps that I'm definitely pushing for, that we can deliver on those visions of protecting the primary storage automatically. For example, it should be automatically protected. So there's a, I think, a great opportunity ahead for us. And I think the competition will have a difficult time because they do not own those assets. They don't own primary storage assets that which play a role in it. They don't own the virtual assets. So I think we can deliver that integration very safely. I have to say, EMC has done a good job of taking the hardware and software and your former company doesn't have that software piece. It's got to partner up. And that always, you know, it's always harder, right? Yeah, exactly, yeah. Okay, Stephen, I'm going to give you the last word. So you said this is the year of virtualization or VMware specifically around data protection. What should we look for in the next six to 12 months in terms of indicators of progress or the milestones that you want to hit? So I think in this space, there's probably three things I really want you looking out for. One is, how easy is it to manage protection from my virtualization interface? You know, one of our profound philosophies is the VM admin doesn't want to log into a backup console. They want to manage everything from tools that they're familiar with, whether it's VCOPs, whether it's the log analytics, whether it's vSphere, whether it's vCloud director. So one, look for all we're doing in terms of being able to really simplify the management from a VM perspective. The second thing I think you want to look for is really, again, the number, the opportunities to be able to bring together backup and disaster recovery. So the ability to handle all kinds of recovery. Full disaster failover, roll back if something gets corrupted, as well as single object recovery. So really bringing together all of that data protection functionality into one unified whole. And then I think the third one you want to look for is, again, our ability to really automate the integration of protection into the environment. And that is the, hey, you know, I suddenly can spin up VMs significantly faster because I don't have to make a separate call to set up protection. I don't have to worry about, is my entire application protected because it's already bundled together and managed. So it'd be those three, native interface management, bring together backup disaster recovery archive into unified data protection and the ability to make everything automated as part of the infrastructure. You'll see those three things over the next six to 12 months. I think it's going to make a huge difference in a lot of these large virtualization environments. Excellent, Stephen, you've really brought a lot of energy and vision to the BRS group. I mean, I think they're, I think they're, first I think they're lucky to have you. So, and we're lucky to have you too. So we really appreciate you coming on theCUBE. You know, we'll be watching. I think it's great that a company with your resources can put forth that vision and can advance it. It takes a large established company, I think, to do that because it's such a hard problem. So thanks again, gentlemen, for coming on theCUBE. Keep it right there, everybody. We're back, Ed Rixas here. He's a CIO. We're going to unpack how he's handling some of these challenges around data protection. This is theCUBE. We'll be right back.