 Everybody, we're back. This is Dave Vellante at Wikibon.org and this is theCUBE and we are drilling into backup and recovery. David Fleuer from Wikibon is with me. Last year at EMC World 2012, Steven Manley, EMC's CTO of the BRS division, put forth what was to us a fresh vision of backup as a service. It's not one that we had heard before from EMC. At Wikibon, we've been pushing for this vision for quite some time, so we got very excited. David Fleuer wrote some posts on it and we're back here a year later. We're going to dig deep into this with Steven Manley, who is the CTO, as I said, of the BRS division. Steven, welcome back. Oh, it's great to see you again, Dave. How are you doing? Good, really great show and really appreciate you guys having us here and all the support that EMC has given us. You guys are fantastic. Nice backdrop, good set. Absolutely. How's the food been? Food's been rushed, you know, because we've been back to back, so I can't really say I tasted it. I've been eating about seven cliff bars a day, so I got nothing. So, Steven, you laid out the vision last year and now it's a year on. EMC now has a sort of a pattern of doing this. Look, this is where we're going. You're very transparent about that. Setting milestones and you guys are pretty good about delivering on them, so we're here a year later. What's new and how have you been executing on that vision? So, yeah, there were a few things that really happened through the course of the year that sort of fit in with the vision, so one of them going kind of chronologically was we had the announcement at VMworld about the vSphere data protection product, and I think we actually talked there as well. And that was really key to the vision because, again, when you start thinking about the data protection value needs to get in the hands of different administrators, really building some of that functionality into vSphere goes a long way towards enabling the VM admins to have a lot more control over their environment. So that was a big step for us. And then, as you fast forward through the year, we had a release of the data protection advisor six product, and data protection advisor really is one of those products that says, look, I really don't care how you're backing up, because a lot of traditional backup applications take that attitude of, if you didn't back it up with my client, I'm not going to touch it. DPA is more in that family of, look, I'm not going to judge how you're backing up, how you're protecting your data, but I want to give you a central visibility into it. So with DPA six, that went a long way for us towards, A, we really simplified the product, so it's a lot easier for people to deploy and roll out, and then we made it a lot more scalable so that you can do not just reporting, but some deeper analytics against the environment. A couple other really cool things that we did this year, we released Data Domain Management Center, which is a product that's really targeted around, especially as you get to larger and larger data domain environments, someone's got to manage those. And one of the things we've heard from a lot of the customers is, this protection storage is a little bit of a different beast than what the rest of our environment looks like. It's not quite like a regular NAS platform, and it's not a SAN platform, so they've in some cases really struggled at who manages it and how. And so with the Data Domain Management Center, it was really to enable the backup team to step up and say, all right, even as maybe our clients and our traditional backup flows start to go away, we're going to add value by managing that infrastructure. So it was really about simplifying, managing a protection storage infrastructure for people that maybe didn't have 20 years of storage management experience. And then really the last piece to get into that we released is a new software version for Data Domain. It's right now in Directed Availability. And a lot of the focus there was really around, A, laying the groundwork for what's going to come out with Avamar soon, but also really increasing our scope in the archival space, because one of the things we look at is it's not just about backup anymore, it's about data protection, which is backup disaster recovery and archive, and so we're really trying to take that platform to cover all those pieces. So Stephen, you heard David and myself speaking earlier about some of the problems with backup. One of the things we didn't touch on that I'd like to address with you is backup as an inhibitor to application development. Can you elaborate on that a little bit? First of all, do you buy that? Oh yeah. And second of all, can you sort of double click on that? Oh absolutely, you know, it's been fascinating to me and I think Guy might have brought up a similar sort of thing this morning, that for a long time we lived in a world where frankly application developers, the ISVs that actually made apps, they didn't worry about backup that much. That was the backup apps problem and it wasn't getting in the way of their business and so whether it was Oracle or SAP or Teradata or even our own Green Plum, backup that's somebody else's problem. And as I've gone around the country for the past couple of years, it's just been case after case after case of a customer saying, I've got a TSM or a DSM, a DB2 database of, yeah, people also. Let's not go there, just tease it, okay? All right, so, sorry, force of mental effort. We don't have that much time. But I've got a DB2 database of X size and there's just no way I can hit a backup window and but people want to keep growing the database because there's a lot of places where the view on their databases is, I never take anything out of the database so it just keeps getting bigger or, you know, again, I had a customer that said, I'm trying to consolidate my exchange servers and the backup window and the recovery window are killing me and so it's becoming the long pole in these application deployments. And so one of the things that really struck us is when Oracle started putting things into Oracle to optimize backup, to let backup applications run faster, that's an enormous shift. I mean, that's the sort of thing that, again, a decade ago you would not have seen an application say, I'm going to do something specifically in my app to enable backup and I think that's really reflective of the industry saying, if I can't get my data backed up and of course if I can't get it recovered, I'm in serious trouble and so they're starting to, frankly, do things to really enable the backup infrastructure to scale. They're doing even more than that. I mean, they're doing investing a lot in Golden Gate and Active Data Guard, they're really putting out a lot in there, which presumably is not, you would like them to be focusing on actually making traditional backups a little bit better. To be honest, it's an interesting dichotomy for us because on the one hand, sure, EMC would love to own as much of the stack as we can. But from my perspective, especially as I look at the evolution of the backup administrator and the value in the backup stack, the value really becomes again in that policy management, the centralized reporting, the cataloging and the infrastructure management. And so I think the backup software vendors that are going to scale and succeed and be around in a decade, they're ones that are really going to have to get off that model of, no, no, let me move the data for you to, all right, you know what, if Data Guard happens to be moving the data and that's the right answer for the customer, I want to respect that, but I also want the protection team to have the visibility they need into it. And so, from my perspective, I want the data to be safe first and foremost, and I want to make sure that as a vendor, we're not saying, no, didn't happen through us, so we refuse to acknowledge it. I want us to be very legitimately saying, no matter how you're protecting your data, we're going to be there for you for that visibility. But the fact that the ISV community has to actually go through those somewhat unnatural acts of a wake-up call for the vendor community, isn't it? Well, I think it is, and again, I think it sets in course, it reflects a course of motion that I think's been going on now for 15 years, which is the traditional backup client architecture long-term doesn't scale. Because the difference between, whether it's my storage array, my application or my hypervisor and a backup client, is that the backup client is not awake and alive during the day, right? It's something that wakes up and then searches for data to back up. But as your data lake gets bigger and bigger, that becomes a bigger and bigger challenge. I contrast that with, if I'm the owner of the data, if I see every bit that's being written through, I can keep track of exactly what I need to back up. And so, again, whether I'm the app doing it, the hypervisor doing it, or the storage doing it, I've gotten advantage that the backup client simply doesn't have. And so, to me, it's not so much that the backup community was asleep at the wheel here, as it can only get you so far. And now, the areas that own the data, they're stepping up to make this possible. And again, the successful backup companies are the ones that say, I accept that that's where the world's going. How can I provide services and add value in those contexts? Okay, so this is how we got to this point of accidental architecture, which Guy Churchwood claims he coined, you stole, by the way. Yes, that's entirely true. He actually coined it at the previous company we were at together. And before he joined here, I had already stolen the phrase, not realizing that I would someday recruit him to come in and do my lawsuit. So, he completely owns that term, but I've said it. So what's needed to see this backup as a service through? So, like I said, I think it's a couple things. And the first and most important one, and I think David was saying the same thing, it's not a technology problem first. It's an orientation and focus of the group first. If you can't fix the people and processes to think in the right way, to think about how they're going to connect to the business and add value, then I don't think this transformation works. And so, a lot of what we do as we transition people away from here's how to optimize the way you used to do it, to here's the new model, is really working with the backup engineers and the data protection engineers and getting them comfortable with, no, we're not taking you out of a job, we're not making you irrelevant. We're actually, frankly, getting rid of a lot of what was an unpleasant part of your job, deploying clients, managing clients, dealing with network outages and getting you into a more high value role. Once you flip that switch and they understand where you want to go, the technology problems become a lot easier because then they connect to the business and say, what is it I need to solve? And then they can start looking for what's the right answer. And whether it's an Avamar solution or data domain or data guard, they're at least looking at it the right way, which isn't, here's what I do, take it or leave it and it's more of, all right, how can I go solve those problems? And as engineers, to be honest, that's a heck of a lot more fun anyway. I like solving problems, not creating them. So can we talk about as well, so you've got all these different assets, many of you bought into the company, how do you tie those together? What, let's talk about them from a component standpoint, what do you have to do to tie those pieces together? And in particular, if I may add something to that, how do you make those resources not tied to a specific application? Make them so that they're usable. All over the place. All over the place and you can extract value from them instead of just leaving them to language. So this is, in a funny way, as I'm up on stage talking to people about the accidental architecture versus the intentional or protection storage architecture, to be honest, we need that internally as much as we help our customers put that in place. Because left to each group's own devices, they probably would say, all right, I'm going to design this for my own little area because my problem, I've got a release to ship, I need to get this out to solve these customers' problems. And I'd wake up with three backup catalogs and three policy engines and five protection storage platforms. So in a lot of ways, that protection storage architecture started as an internal design. And of course, you double click and you can get to a lot more specificity so that the engineers as they design it understand the context that they're working in as well. So just like our customers need an intentional architecture, we needed one too. And that really guides you towards, all right guys, remember, it's not even just Avamar Network or Backup Software, it's Viper, it's VMware, it's Oracle, it's SAP. You need to work in all of these environments. And again, engineers just want to solve problems. If you put the context and help them understand what they need to do, they do the right thing. These are smart people we're dealing with. When you leave them in the dark, that's when bad things happen. So, Steven, you talked about Viper just now. And Viper's vision and part of the vision is you can run it on anything, right? But you talk about this protection storage architecture. Why do you need a protection storage architecture versus just running it on anything? Right, that's a great question. Because again, when I look at Viper, the best part of Viper is that it really takes all these purpose built arrays in a lot of ways, right? Whether it's the data domain, the VMAX, the Icelon, and says each of them are really good at one thing. And the more we try to make every, each platform do everything, the less you lose your value and the more you become a one size fits all. And so, when I look at each of these design centers, protection storage in particular, I look at the data domain platform and it's different than really any other platform designed on the market, at least at this scale. And it's because, A, it was built from the ground up to be for data durability. So, there's this data and vulnerability architecture we talk about where we've got levels of check sums and reference checks and validations that no other platform has. Now, I'll be honest, these don't come without a cost. There's a reason a data domain doesn't run an OLTP database like a VMAX does because we have these check sums and these validations. But it's built for that. The second thing that it's built for is at a fundamental level, block sharing. And so, a lot of people, oh no, it's built for DDoop. Well, it was built first and foremost to share data. And what that lets you do is whether it's, you're sharing it because you're deduplicating it or you're sharing it because you're getting these change block tracking backups from VMware or from Oracle or from EMC primary storage, it lets me store each applications back up individually with their own retention period in their own space without getting into complexity as a management. And then on top of that, you get the deduplication, the compression and all the space efficiency. So, to me, there's no other platform in the market at this scale like that because all the other platforms had a different design center in mind. And yeah, you could use another platform but you start getting into that one size fits all of, you know, you're not getting the best result that you could possibly get. All right, Steve, we're almost out of time but I wanted to give you one last question. So, as observers, customers, analysts, whomever press, watching you move down this journey, you put forth the vision last year, now you're executing it, what are the things that we should watch for? What are the milestones? How are you judging yourself and how should we judge you? So, I think if I were looking for the big milestones, I break them into two categories. One, sort of the short term, go to market, visible things and then the other on the kind of the technology coolness. So, I wanted to do the technology one first because that's more fun. So, on the technology, first and foremost, you should see out of us over the coming 24 months in incremental releases, really an entire re-imagination of data protection cataloging. At a scale that no one's seen, with the flexibility that no one's seen, that, again, is not directly coupled to any specific application that should be able to span from, like I said in the keynote, from local snapshots all the way up to cloud versions. And for me, technologically, that is the single hottest, most fun, new, sexy technology that we're working on. And to me, if we nail that, that's really transformative for, again, the role of the backup team because now they've got the crown jewels where they know everything. So that one gets me real fired up. The other technology gets me fired up is, again, the data domain moving up the stack from, I don't just hold backups to, I've got backups and archives to backup archive, ooh, and you can start to do a little DR off of me. You can start to spin up VMs and applications and run them off of me, not as fast as the primary, but better than not being up. On the go-to-market side, I think you want to look towards a couple things. You want to look towards, again, more and more bringing together the backup software, the Avamar Networker. Guy talked about the data protection suite, but I want to follow the data protection suite with actual, again, deeper technical integration. I think the other one that you want to look at is, especially over the next year or so, more and more, what's the primary storage and the protection storage at EMC? How are they integrating? How are we delivering in that part of the Viper vision? Because that, you're going to see some things. Excellent, all right, Steven. So is that going to allow you to actually free the data and so that you can do it, analyze it? Ah, that's exciting. All right, Steven Manley, thanks very much for coming on theCUBE and helping us unpack this exciting vision and starting to see it come to fruition. So congratulations on the progress that you've made so far and appreciate you laying out those milestones in a very transparent way, so we'll be watching. I was a pleasure to talk to you guys. All right, keep it right there. Everybody will be back with Federal Express to dig deeper into how they are dealing with their backup challenges. This is theCUBE, we'll be right back.