 From Las Vegas, it's theCUBE. Covering VMworld 2018. Brought to you by VMware and its ecosystem partners. Welcome back to Las Vegas, everybody. This is theCUBE, the leader in live tech coverage. We go out to the events, we extract the signal from the noise. This is VMworld 2018, our ninth year of VMworld at theCUBE, we got two sets here. VMworld, unbelievable, they built out the sets. We have great guests, 94 guests this week. I'm Dave Vellante with my co-host, David Floyer. Emmett Cosmeric is here. He's the field CTO for data protection sales at Dell EMC, he's joined by Christoph Bertrand. Good friend who's an analyst now at ESG. Gentlemen, great to see you. Thanks for coming on theCUBE. Thanks for having us. You're welcome, Emmett, let's start with you. I mean, this has been like two years in a row now. Data protection is the hottest topic. Cloud, things evolving from just a virtualized world to this multi-cloud world. People rethinking their whole data protection practices. What are you seeing? Give us the update. Well, absolutely. I mean, there's a huge focus as a whole now around data protection. You've got rising things like ransomware and attacks out there that people are really evaluating. Are they truly protected? You're also now seeing the cloud being leveraged more and more as a way to integrate in with your backups, using it for long-term attention, using it as a target for DR. So what's really interesting from that adoption rate, if you want to actually look at where we were a year ago, specifically with AWS, for example, we were driving three petabytes a month for consumption with our data protection products into the Amazon cloud. Today, that's up to 20 petabytes a month. So we're seeing a massive adoption by our customers in that market there. And what's really interesting when you start to actually double click on that, when you look at the efficiencies we have provided for our customers with our deduplication technology, by shrinking that total footprint down that they actually consume, there's a lot more economics of scale that they get to benefit from as a result of that. When you talk about going from three petabytes to 20 petabytes, you're talking about using the cloud as a data protection target, right? Absolutely, long-term retention or now with our cloud DR capabilities and the ability for customers to actually fail over and burst into the cloud. Another prime example of where we're driving that up. Another example as well is protecting workloads in the cloud. Recently, we were able to launch and customers can deploy through either the AWS Marketplace or the Azure Marketplace's data domain virtual on object storage. So not only are you getting the benefit of that market-leading protection capability of data domain as a software-defined instance, but you now have a cloud-centric running on object storage. So even more cost efficiencies that it's bringing for our customers. So Christophe, what is your research showing in terms of adoption of cloud in the data protection context, whether it's as a target or what do your research show? Yeah, absolutely. AWS, we know from our respondents to our surveys, have doubled in usage in the past four years. So it's crazy, right? So definitely a lot of adoption. As far as data protection goes, a couple of different perspectives. One of them is, of course, the quest to always reduce costs as a major IT spending initiative and the fact that cloud can help but can also be, if you don't pay attention, an area of additional costs. So you tie that back to data protection, definitely being used more as a target, as a natural extension to your on-premise environment, very naturally. But also, in some cases, as cloud-native, as cloud-native workloads evolve, well, there are some changes in how you protect data. So lots of evolution. I think we're really in the world now of hybrid cloud, for sure, and hybrid data protection, which is why you need to really think about how do you extend the great capabilities you built on-premise and make them work in a cloud environment? Data deduplication is one of the big areas that I've always been very interested in, because I've always been told, follow the money. Well, the money is where the storage, where there's compute. And the less you use, the less you're going to pay. And I think that's really what you have to look at as your deploying solutions that leverage cloud. How do you optimize the spend and how do you make sure that you meet your S&As? Let's talk about the ROI, a simple equation, benefit over cost. And traditionally, the CFO would look at the benefit of data protection as, I got to do this, because my CIO said, I have to, or we'll lose data. I said, all right, go ahead. So there's been a huge focus on the denominator, and still is. Is that changing, is the value component, the numerator changing in any way, shape, or form as a result of cloud with maybe new use cases beyond just back up and restore, or is it still kind of hardcore, reduce the cost, in thoughts, Emmett? Well, I think it still is always going to come down to dollars and cents, right? That is a huge aspect of what we're trying to do and accomplish across the board as organizations and what our customers are looking to do. If they can drive down that total cost, use less resources, it's a good thing for them. However, do you have drivers around compliance? They have drivers around needs there. And what we're seeing is the cloud allows us to do some very interesting things. Cheap and deep storage and object allows us to have some really interesting use cases there, and you layer on top of that then, deduplication and using and deploying less resources as a whole to protect that. There's added benefits across the board. What also gets very interesting now is we're at VMworld. So when we're talking about VMC on AWS, we can use these same services and we can use S3 as that object bucket from that native Amazon when in VMworld at cloud. So being able to take those protection capabilities that our customers were leveraging on-prem with our proven solutions, deploy them up in VMC, and still leverage also that power of having that S3 bucket on top of deduplication with it, equates the less cost for the customer and more efficiencies overall. So can you give some examples of how you can exploit that data in the cloud, do more with it than just have it there to recover from? Is that an area that's focused for you or can you give some examples? So absolutely, one of the areas we're starting to see more and more is customers wanting to take data and replicate it up in the cloud and then redeploy it up there. So you look at database instances and their capabilities of having that native integration we can do with Oracle, and then being able to then take that RMAN backups and redeploy it as an RMAN instance up in AWS or up in Azure, it makes things very interesting for customers on how they want to deploy it holistically. The other side of this is, it comes down more to how do I want to leverage the cloud? Do I want to use it for long-term retention to eliminate tape? Do I want to use it as an N plus one target? Do I want to use it as an alternative option to export DR, to actually burst in and then pay for as I need it, by the drip, those resources when I need to have that DR event. But one of the key aspects there that we really focus in on, it's not just about failing in, it's about failing back because our customers have made this investment on-prem. They want to come back eventually. So one of the key aspects with our technology is we have orchestrated fail-back as part of the process to help them get back on-prem when it's appropriate. Chris, are there hidden costs that customers should be thinking about with regard to data protection and what are those? Well, so if you look at data protection in general, and I'll just start from the S&A perspective. First of all, it's important to remind people that about 15% at least of the people we surveyed, which are traditional mid-market and enterprise organizations, do not want to see any downtime, okay? So that's the first thing. And then you go to the one-hour bar and you're probably at 60, 70%, right? So that's the first statement because you've got to have to pay at some point for that S&A. Now, you look at cloud more specifically and how you can optimize the costs. I mentioned data duplication. As a matter of fact, we've validated that the data EMC technology can limit the cost significantly when using either the ADPA appliance or the combination of software solutions. We have a couple of reports on that. If we look at a survey of about, we're looking at telemetry from about 15,000 customers, we were somewhere between 51% to 81% of costs savings, meaning the technology has really helped. Why? Compute is one issue to answer your question. So how do you find a way with the technology to limit the amount of compute so that the virtual machines or whatever servers you use in the cloud do not use up too much of that compute? The other is, of course, storage. In the case of data EMC, I think we've identified close to 71, 72% savings of S3 storage because of the grant deduplication that's built in. So these are the costs that we look at, look at, again, look at the bill, look at the ingress, the egress, and look at what you're consuming and where can you apply technology that's going to still allow you to deliver on the SLAs, never forget the SLAs, pay a little more if you have to, because when you talk about the cost of data protection, there's also the cost of not protecting data and that can be much higher. But look at those dimensions and you'll do well and that's where the technology kicks in. So to just kind of double click on something that Chris Hofstra said there, their findings at ESG showed that we are saving, on average, 73% on S3 costs for our customers when they leverage our technology. You think of the fact that we, on a monthly basis, right now are driving 20 petabytes of consumption in the cloud. That 73% savings, what would that look like if it wasn't there? So that is real dollars and cents that these customers, if they weren't leveraging these technologies, would be forced to actually be paying out of their pocket. What are the customers actually deploying? What's the solution look like that allows them to get that 73% savings? So it's going to come down to what's the use case we're trying to solve, right? So I mentioned we have the capability of doing cloud DR, to burst in from that aspect. We have the ability to actually leverage the cloud for long-term retention, using it as a target, either via the data domain for cloud tiering itself or if we want to just send it right directly out to object storage in the cloud, we can use our native built-in capabilities to do that as well. And then finally, protecting workloads in the cloud. We can deploy our technology as software to find instances through the marketplaces and run it on object storage. So that's really allowing customers to have the options of what's the use case we're trying to solve and making sure they're using the right tool to do it. Great. Are you getting some demand for people who want to actually back up on their own site, applications running in the cloud itself? Can you support that? That's a great question. So what we're seeing in certain areas, they want to do replication back. They want that on-prem to be that N plus one and have that there. And so it's a great area of when using data domain virtual in the cloud and replicating back to an IDPA or our traditional data domain appliances allows them that capability. And we still give them the ability to have a central data management plane across the board. Doesn't matter if they're deployed solutions from a remote side aspect, from their data centers or up in multi-clouds, they have that one visibility in view. Excellent. Christof, what are you seeing in terms of the big gaps between what the customers are expecting in terms of the data protection experience and what the industry has traditionally been able to deliver? And then of course, Emmett, I'm going to ask you how you're closing those gaps. So think about that. So there are a few changes. Some may have been part of the industry as an actor for many, many years. So now being on the other side as an analyst, it's very interesting to see is it from a slightly different prism. I will say, first of all, I would call this the SLA gap. Was mentioning those trends about what is really acceptable. Well, the real question is, do you really deliver against those SLAs? Interestingly, it's not just for mission critical data. Our research indicates that even non-mission critical data has become very sensitive to SLAs, meaning data loss and downtime to keep it simple from a data protection standpoint. So I would say that's the first gap. Do you have, do you meet those SLAs, at least for your critical applications and then do you meet it overall? The other thing is there is a lack of coherency across the various applications and workloads you're trying to protect. So having a solution that is as broad as possible in many ways, or at least having the processes and the technologies in place that are coordinated enough that you're going to have that coherency is going to be very important because guess what happens when something happens and typically in a data center it's going to be some sort of hardware failure or human error, you name it, or a cyber attack, big, big thing these days. Well, you're going to fail over to a cloud environment, you'd better have all of your applications come back in exactly the right order, you'd better have that run book clearly worked out. So I think that could be the second areas of gaps as far as having your backups, having recent backups and high availability or near high availability. Do you have the processes? Do you know what to bring back in what sequence? So any solution that's going to give you the ability to orchestrate is going to be very key. So I think there are gaps still today but the good news is there's a lot of technology that's happening here. So we're seeing it from Dell EMC with different ways to consume it, whether you want to consume software or hardware combination thereof or appliances or converge or hyper-converged systems. So lots of options. So the gaps are closing, they're still there and guess what, 10 years from now we'll still be talking about data protection. Hey, no doubt, that's great analysis. So Emmett, it's consistent with what you're hearing from customers and give us a little sort of glimpse as to how you're closing that gaps. No, absolutely. So you see a range from customers. You see the ones that we just want to go to cloud. Okay, well what does that mean to you? I don't know, but we need to go to cloud. And then you got the ones who are actually going through the process of actually cloud architecting and going and looking at the applications and determining how they want to do it. And then one of the things I love about VMC and VMware as a whole now in AWS is it's being able to bring that power of that trusted performance of VMware up into that cloud experience so that customers are able to use the tools that they've been comfortable with that they've designed with that they trust to now also run in the public cloud. And what we've done specifically in that area is we were one of the first manufacturers and vendors out there to support protecting workloads in VMC. And so continuing on what Michael's mission is and working together closely under the Dell Technologies family to make sure we're coherently integrating in and bringing more protection capabilities for what our customers need up there as a whole. Shifting back though slightly, looking at public cloud just across the board. It's about making sure that we're hitting those gaps, that orchestration on the failback. There's been the capability for people to fail into the cloud. I mean, Amazon released a tool to do that conversion and cut over. But the gap that existed there was the failback. The other gap that existed is there are certain workloads that just aren't suitable to run on AWS. Linux instance VMs that are over two terabytes. They just won't work. So making sure that we notify customers when they try and select something like that on the front end that, hey, this isn't a supported workload. So they're not trying in a DR scenario to be like, why is this not coming up? So putting those pieces in place for our customers so that they have the ability to know on the front end and also have that workflow back has been one of the really key areas we've been focused on. Awesome, Emmett, thanks so much for sort of bridging the world of technology and customer. We really appreciate the explanation. Kristoff, great analysis. I always appreciate having ESG on. You guys do great work. So thank you guys for coming on. Absolutely. All right, keep it right there. Everybody will be back day three VMworld 2018. You're watching theCUBE. We'll be right back.