 The Cube at EMC World 2014 is brought to you by EMC. Redefine VCE, innovating the world's first converged infrastructure solution for private cloud computing. Brocade, say goodbye to the status quo and hello to Brocade. We're back, welcome to Las Vegas everybody. This is Dave Vellante with my co-host Steve Keniston, my long-time co-host, a co-host of Running Data, many of you remember that program. We're here live at EMC World 2014. Our good friend Jason Buffington is here. This is the detailed spotlight on data protection, something that we've done with Jason now a few times. Jason, welcome back to theCUBE, it's great to see you again. Thanks for having me. Yeah, so EMC World 2014 keeps getting bigger. It does. And I think that means better. Yeah. What's the number? Have you guys heard the number here? It's 12,000 anywhere between 12 and 50. So, you know, that's fluffier. That's pretty good. You know, a little data protection love. I mean, for a while there, post-acquisition of data domain, it's all EMCs to talk about now. It's sort of getting buried inside of flash and in cloud and a box, et cetera. But ever since Jason and I've been in this business, data protection has been the number one challenge for IT practitioners. How do I protect my data? How do I hit my backup windows? What happens when I lose data? Is that conversation still going on with customers? It is absolutely still going on. So, every year ESG asks about what do you plan on spending money on? Every analyst firm has their own way of figuring out which way the wind is blowing. We ask people, what are you going to pay for? What are you actually going to spend money on this year? And data protection's always in the top three. In fact, for the last three years running, data protection's actually been the number one planned spend for mid-sized organizations year over year. One of the things I think what's interesting is, if the only thing you were trying to solve for was I need to backup my servers, it wouldn't be a top three. It wouldn't be a top 10. Everybody's got that figured out. The problem is that the workloads keep evolving. And because of that, we have to reinvent backup. So, virtualization has never been more prolific than it is now, and that's changing how we do backup. You got workloads going to the cloud. The cloud wasn't there five years ago. For all these different workloads, as the workloads evolve, we have to rediscover or reinvent how to back it all up. And that's why it's still top of mind year over year. Okay, so I'm picturing just a number of different ways to do backup in not just one. First of all, we've talked about one size doesn't fit all. We don't have to go back to that. We can, but what I'm thinking about is just multiple different techniques that are applied to different parts of your infrastructure. So, paint that picture for us. What's that spectrum look like? And spectrum's the right word. It's actually what we call it. We talk about it as colors of a rainbow almost. So think about this. It wasn't that many years ago that the marketing people would confuse the terms data protection and backup as being synonymous. They're not. Backup is a very small subset of a data protection spectrum. In a logo on my blog, I think backup is red. But the other colors in that spectrum, you've got archiving, which is different than backup. You've got snapshotting, replication, availability technologies, business continuity, disaster recovery. There's seven different kinds of things that you could call part of that data protection discipline. And each one, by the way, it's stone color. So the conversation that we're trying to help IT pros as well as vendors have is when was the last time you saw a rainbow that didn't have green or they didn't have blue, right? Your data protection strategy, your spectrum really ought to have all the colors of that rainbow in it until you figure out otherwise. And to your point, sometimes it's based on workload. What I would tell anyone who cares to listen is start first with what the business goals are of what you need to be able to recover. Is it over distance? Is it immediate failback? Is it previous versions? And then from there, that points you to a color. And then from there, you start looking at products. People got it backwards. So that's an RTO and RPO conversation. Or maybe some other parlance, however you communicate to the business, right Steve? We've talked about this a lot. Maybe RTO and RPO is something that turns off business people. But we know what we're talking about. How fast do you want your data back? How much data you're willing to lose and how much money do you have to spend? And we've talked a lot about the fact that, that whole term, that definition I would agree Jeff, that definition of data protection and backup kind of being synonymous with what happens. But you painted a very interesting picture along that spectrum. So I think the pendulum swings both ways when we talk about, okay, there is no one size fits all. There's no one application that's going to do it. But now we're hearing grumblings of a couple of things, right? We're hearing things like copy data services, which might be able to do some of, more than just one of those things. And then we're also hearing the fact that it's not just about the backup now. It's about getting that backup application closer to the application. How do you see that dynamic? Yeah, absolutely. So let me do those backwards. As far as the, how do we get the data protection closer to the workload could not agree more. And that's not only the, how do I get recovery and that protection closer to the workloads. Also things like, how do I get DDoP closer to that workload? So whatever you can do to be more intimate and more aware or savvy of the application, I think the better off you are. One thing I want to make sure that we're clear on though is just because we're talking about seven colors of a rainbow, you should not settle for seven or more different products, each one just satisfying a niche along the way. At a minimum, you should be thinking about management UIs, you should be thinking about protection silos of storage that actually let you accomplish multiple solution sets. You shouldn't have to niche through those because even if you had seven different colors, by the way, in backup, you could have two or three different backup products just in that category, who wants to manage 10 UIs, right? That's insane. So yeah, so we want to make sure that we're talking about recovery goals and strategies that doesn't necessarily match up to a hodge podge of unrelated products and silos. And what's the biggest thing you're seeing when it comes to this hodge podge? People, you know, we've been saying backup is broken for a while. You know, it's been a term. Do I think it's broken still? Maybe, but to your point, we've seen this evolution of workloads come through and it's kind of like everybody keeps saying, well, as they evolve, maybe I'll start to grow into my new protection solution and then it doesn't fully evolve and now you've got two or then you've got three and then you've got a workload that might need CDP. So now you've got a bunch of different things. So what are you seeing? So I guess the person I'd say is it's not that backup is broken. Just stop trying to use a hammer as a screwdriver, right? So backup is one color of that spectrum, right? But it doesn't mean it's a one size fits all deal. And so when people talk about backup being broken, what they mean is the legacy gadget that I've been using forever no longer keeps up with the workloads that have evolved. Well, shame on you, if your workloads have evolved and you haven't evolved your protection in order at the same rate that you've evolved your production, then yeah, you're going to outgrow that. It's going to be broken. What it really needs to be is look, start looking at what your production workloads are and what their different needs are and what you're going to find is that the protection schemes need to change. So if you start going through, say a highly virtualized environment, you can't keep putting agents inside of VMs and getting that done. You're going to need to be looking for a host base. You're going to need to start thinking about snapshotting and rollback. You're going to need to start looking at protection. Let's do instant VM recovery. But as the production workload evolves, protection infrastructure has to evolve with it or you're going to be broken. So you make a good point, right? So I want to go back to something you said before. So if for the last three years, the industry has said that one of the number one places where they're actually spending money is data protection. And you've got this primary storage workload that's evolving, right? But if people are spending money on backup and they're not keeping up, what's the gap? What are we missing? That's a great question. So I would say the gap comes in twofold. One, there was a gap in the quality of virtualization protection that frankly, all the legacy guys created what, five, six years ago, right? So all the big guys out there that were backing up everything on the planet, they weren't as agile with VM based protection as they should have been when VMware was first really taking off. And so it created a niche where you've got these virtualization specific backup solutions that are coming into the fray. Now, we've got a scenario where the unified solutions want to get that market share back. So they are trying to catch up on parity of features. The virtualization specific backup solutions, they're still continuing to innovate out front in many cases. And then oh by the way, just to keep things really interesting, VMware is now shipping VDP. So you have to compete with backup which is now in the box. It's a really weird time where people are trying to figure out how you put this all back together again. And so seeing things like what we saw at EMC just a few months ago where they announced where Avamar and VDP are now starting to talk to each other, that you can start to see now. We get that convergence back again. And oh by the way, VDP is now talking to AdoMain. And so we're seeing it get simple again. But for a while it was. So I might have been spending money. I was spending money on my traditional apps. The environment was evolving. I should have been spending on smarter things but the legacy things weren't keeping up. Well my gripe on backup is broken. There's always been, generally you walk into an environment and you say how do you do backup? And they have one way of doing it. And it's broken. It is a one size fits all sort of problem. And maybe if they have a lot of money they could do some kind of high end replication deal. But generally speaking it's daily incremental, weekly full, no matter what your RTO and RPO. So either you're spending too much or you're spending too little or maybe there's an app where it's just right. So the point is this notion of a service catalog we talk about backup as a service a lot. Is that something that is becoming a reality in some of the customers that you talk to? Yeah it absolutely is. Especially when you find a workload that is so different than what you're used to doing you truly just cannot use what you used to use before. So for example when you think about virtualization but I'm going to get off that in a second. When you're 20% virtualized you can pretty much use any mediocre solution you want. When you get to 40%, 50%, 60% virtualized not only the legacy approach does not work they can actually hinder production, right? Because they start taking head room that you don't have in a highly converged environment. But start thinking about some of the newer workloads where people are finally realizing that they got to protect data like the cloud. When we talk about backing up and cloud most people want to talk about backing up to the cloud. One of the big problems people are just now starting to realize is what about backing up of the cloud? You know you move that workload onto a software as a service-based platform you don't get the luxury of using what you did before even if you wanted to, even if you were lazy about it because that stuff doesn't go to the cloud, right? So all of a sudden you've got to reinvent that. The same thing holds true for like end point scenarios, right? Tell any IT administrator hey the way that you've been backing up these 400 servers I now need you to do that for 20,000 desktops. Nobody's doing that, right? So all of a sudden we start looking at okay we'll back up as a service starts to become interesting at that point. Merging, Baz and online file sharing starts to become interesting at that point. Same thing for SaaS by the way if you're going to back up something that's in the cloud you should back it up with a service that's also in the cloud as well. Why would you pull that back again? So as the workloads change to where you don't even have that option it forces people to add, not replace their data protection tool set and that's where that service catalog has to come from. What about you brought up the SaaSification of IT? What are the backup considerations or let's say you're a whatever a work day or a service now client? I mean it's your data. What are the backup considerations for those types of SaaS products and how do I ensure consistency with the edicts of my organization and check the compliance boxes? There is a whole bunch in that question which is totally unsolved. So one of the challenges this has been true for as long as I've been in the industry probably you as well. You know whenever a new platform really kind of takes off the idea of the plumbing for enabling backup it's never in the V1 or the V2 of that platform. It's not until it really comes a mainstay solution that people say oh yeah I really ought to put the plumbing in there to back that stuff up. Think about things like Salesforce. Everything from the biggest of companies to small SMBs are using that. There's no plumbing for backup in that, right? Now it'll eventually get there but up until then you're going to have a very few innovators that are saying we need to solve that problem no differently than the VM guys who were backing it up not five or six years ago. The VMs, the PhDs of the world and before that when it was Windows and you know you've always had these few folks that tried to solve it first and the rest of the industry finally catches up but yeah when the cloud backups are going out there there is no plumbing for that yet. So my last question, we're up against the clock here. Last week we had Frank Slutman of theCUBE and he said to say hi to all his friends at EMC and the reason I bring that up is because the purpose built backup appliance market is now bigger than the tape market that it ostensibly replaced. Does that surprise you? It doesn't surprise me. Now I will say, I'm not a big fan of the term purpose built backup appliance, you know? Yeah, it's those other guys that do that too, right, right. How about third platform? No, no, no, it's not that. It's just that, here's the challenge. You really have to break that category up into two. There is those folks that have disk based target systems. Okay. Okay, data domain is an example, right? But then you also have that pre-canned hardware plus software solution powered up and let it ride, Avamar style, right? So when you put those two together, it kind of confuses the market. And they're solving a similar problem just for different use cases, right? But the difference is that for data domain you still have to have some kind of backup software feed it. That's the problem. It's not an integrated Avamar. Like an Avamar type solution. But to your point, no, that doesn't surprise me at all that we've seen that a SERP tape. I would say this though, the same way that I would say that every data protection strategy should have all the colors of that rainbow, you should start every planning session presuming that you are going to be hybrid with tape plus cloud plus disk. And that disk really needs to be deduced. Tape is not dead. Tape is not dead. I'm glad you said tape is not dead. Plan for all three until you talk yourself out. It's not exciting. It's just not dead. It's not exciting. But it is required. When was the last time you saw a cloud solution that wanted to store your data for 25 years? And since they haven't been clouded in 25 years, it'd be really hard to press the food. Okay, we can't end on tape. So what are you looking for in backup for the next 12 to 24 months? You know what answer I want? How do you backup a petabyte of data? We've had that question, right Dave? How do you back data? I don't, maybe you don't. No, and when you talk about big data, actually you don't. It's actually sometimes, more often than not, it's easier to re-ingest and reprocess that data than it is to try to restore that data farm to an older set. Right. Okay, give you the last word. What should we be looking for? Next 12 to 24 months? First 12, 24 months, what I would tell you is, is that you need to grow past just backup. It's got to be that spectrum-based approach. And more and more often what we're seeing is, is that people really are to a point where they can say, look, I want to start with a recovery strategy, work my way backward into those colors. And then, but once I know what colors, what methods I'm using, fewer silos, better management, and better economics. That's what we should see over the next 24 months. Jason, you're awesome. DuPlessie and Prigmore are lucky to have you. Thanks for coming to theCUBE. It's always a pleasure to see you. Keep it right there, everybody. We'll be back with our next segment in the data protection spotlight. We're live from EMC World and this is theCUBE.