>> live from San Francisco, celebrating 10 years of high tech coverage. It's the Cube covering Veum World 2019. Brought to you by VM Wear and its ecosystem partners. >> Welcome back to San Francisco, everybody. My name is David Lantz. I'm here with my co host John Troia. This is Day three of V M World 2019 2 sets. >> This is >> our 10th year at the M. World Cube is the leader in live enterprise tech coverage. Marry on Blondie is here. He's the C m o and chief evangelist that 10 tree by DDN Yes, sir. He's joined by Graham Breezes The Field CTO at 10 Tree also by DDN Recent acquisition jets Great to see you. >> Likewise, as they say, we're back. I like I like to call it a hibernation in the sense that people may have not known where did Ian or 10 Trias and Tension by Dede and, as the name implies, were acquired a year ago at the M World August 31st of 2018. And in the year since, we've been ableto invest in engineering support, my joining the company in marketing to take this solution, we've been able to save thousands of customers millions of man hours and bring it to a larger number of users. Way >> first saw 10 tree, we said, Wow, this is all about simplification. And Jonah Course you remember that when you go back to the early early Dick Cube days of of'em World, very complex storage was a major challenge. 10 Tree was all about simplifying that. Of course, we know DDN as well is the high performance specialist and have worked with those guys for a number of years. But take >> us >> back Married to the original vision of 10 Cherie. Is that original vision still alive? How was it evolved? >> Well, I'd say that it's, ah, number one reason why we're a part of the DD and family of brands because, as, ah, portfolio company, they're looking good. Bring technologies. I'm the marketing guy for our enterprise or virtual ization audience, and the product sets that cover high performance computing have their own audience. So for me, I'm focused on that. Graham's also focused on that, and, uh, really what continues to make us different today is the fact we were designed to learn from the beginning to understand how virtual machines end to end work with infrastructure. And that's really the foundation of what makes us different today. The same thing, right? >> So from the very beginning we were we were built to understand the work clothes that we service in the data center. So and that was virtual machines. We service those on multiple hyper visors today in terms of being able to understand those workloads intrinsically gives us a tremendous capability. Thio place. I owe again understanding that the infrastructure network storage, hyper visor, uh, weaken view that end end in terms of a latent a graph and give customers and insight into the infrastructure how it's performing. I would say that we're actually extending that further ways in terms of additional workload that we're gonna be able to take on later this year. >> So I know a lot >> of storage admits, although I I only play one on >> TV, but, uh, no, consistently >> throughout the years, right? 10 tree user experiences that is the forefront there. And in fact, they they often some people have said, You know what? I really want to get something done. I grab my tent Reeboks and so it can't talk. Maybe some examples of one example of why the user experience how the user experiences differ or why, why it's different. >> I'll start off by saying that I had a chance being new to the company just two weeks to meet a lot of 10 tree users. And prior to taking the job, I talkto us some folks behind the scenes, and they all told me the same thing. But what I was so interested to hear is that if they didn't have 10 tree, they'd otherwise not have the time to do the automation work, the research work, the strategy work or even the firefighting that's vital to their everyday operations. Right? So it's like, of course, I don't need to manage it. If I did, I wouldn't be able to do all these other things. And I think that's it. Rings true right that it's hard to quantify that time savings because people say, 0 1/2 of it. See, that's really not much of the greater scheme of things. I don't know. 1/2 50. Working on strategic program is a huge opportunity. Let's see >> the value of 10 tree to our end users and we've heard from a lot of them this week actually spent a fantastic event hearing from many of our passionate consumers. From the very beginning. We wanted to build a product that ultimately customers care about, and we've seen that this week in droves. But I would say the going back to what they get out of it. It's the values and what they don't have to do, so they don't have to carve up ones. They don't have to carve up volumes. All they have to do is work with the units of infrastructure that air native to their environment, v ems. They deal with everything in their environment from our virtual machine perspective, virtual machines, one thing across the infrastructure. Again, they can add those virtual machines seamlessly. They can add those in seconds they don't have toe size and add anything in terms of how am I gonna divide up the storage coming in a provisional I Oh, how am I going to get the technical pieces right? Uh, they basically just get place v EMS, and we have a very simplistic way to give them Ah, visualization into that because we understand that virtual machine and what it takes to service. It comes right back to them in terms of time savings that are tremendous in terms of that. >> So let's deal with the elephant in the room. So, so 10 tree. We've talked about all the great stuff in the original founding vision. But then I ran into some troubles, right? And so what? How do you deal with that with customers in terms of just their perception of what what occurred you guys did the eye poets, et cetera, take us through how you're making sure customers are cool with you guys. >> I'm naturally, glass is half full kind of guy from previous, uh, times on the Cube. The interesting thing is, not a lot of people actually knew. Maybe we didn't create enough brand recognition in the past for people to even know that there was a transition. There were even some of our customers. And Graham, you can pile on this that because they don't manage the product every day because they don't have to. It's kind of so easy they even for gotten a lot about it on don't spend a lot of time. I'd say that the reason why we are able to continue. Invest today a year after the acquisition is because retaining existing customers was something that was very successful, and to a lot of them, you can add comments. It wasn't easy to switch to something. They could just switch to something else because there's no other product, does these automatic things and provides the predictive modeling that they're used to. So it's like what we switched to so they just kept going, and to them, they've given us a lot of great feedback. Being owned by the largest private storage company on planet Earth has the advantages of strong source of supply. Great Leverett reverse logistics partnerships with suppliers as a bigger company to be able to service them. Long >> trial wasn't broke, so you didn't need to fix it. And you were ableto maintain obviously a large portion of that customer base. And what was this service experience like? And how is that evolving? And what is Dede and bring to the table? >> So, uh, boy DD and brings so many resources in terms of bringing this from the point when they bought us last year. A year ago today, I think we transition with about 40 people in the company. We're up about 200 now, so Ah, serious investment. Obviously, that's ah have been a pretty heavy job in terms of building that thing back up. Uh, service and support we've put all of the resource is the stated goal coming across the acquisition was they have, ah, 10. Tree support tender by DNC would be better than where 10 tree support was. We fought them on >> rate scores, too. So it's hard to go from there. Right? And >> I would say what we've been doing on that today. I mean, in terms of the S L. A's, I think those were as good as they've ever been from that perspective. So we have a big team behind us that are working really hard to make sure that the customer experience is exactly what we want. A 10 tree experience to be >> So big messages at this This show, of course, multi cloud kubernetes solving climate change, fixing the homeless problem in San Francisco. I'm not hearing that from you guys. What's what's your key message to the VM world? >> Well, I personally believe that there's a lot of opportunity to invest in improving operations that are already pretty darn stable, operating these environments, talking to folks here on the floor. These new technologies you're talking about are certainly gonna change the way we deploy things. But there's gonna be a lot of time left Still operating virtualized server infrastructure and accelerating VD I deployments to just operationalized things better. We're hoping that folks choose some new technologies out there. I mean, there's a bill was a lot of hype in past years. About what technology to choose. We're all flash infrastructure, but well, I'd liketo for the say were intelligent infrastructure. We have 10 and 40 get boards were all flash, but that's not what you choose this. You choose this because you're able to take their operations and spend more your time on the apse because you're not messing around with that low level infrastructure. I think that there's a renaissance of, of, of investment and opportunity to innovate in that space into Graham's point about going further up the stack. We now have data database technology that we can show gives database administrators the direct ability to self service their own cloning, their own, staging their own operations, which otherwise would be a complex set of trouble tickets internally to provision the environment. Everyone loves to self service. That's really big. I think our customers love. It's a self service aspect. I see the self service and >> the ability to d'oh again, not have to worry about all the things that they don't have to do in terms of again not having to get into those details. A cz Morrow mentioned in terms of the database side, that's, ah, workload, the workload intelligence that we've already had for virtual machines. We can now service that database object natively. We're going to do sequel server later this year, uh, being ableto again, being able to see where whether or not they've got a host or a network or a storage problem being able to see where those the that unit they're serving, having that inside is tremendously powerful. Also being able the snapshot to be able to clone to be able thio manage and protect that database in a native way. Not having to worry about, you know, going into a console, worrying about the underlying every structure, the ones, the volumes, all the pieces that might people people would have to get involved with maybe moving from, like, production to test and those kinds of things. So it's the simplicity, eyes all the things that you really don't have to do across the getting down in terms of one's the volumes, the sizing exercises one of our customers put it. Best thing. You know, I hear a lot of things back from different customer. If he says the country, the sentry box is the best employee has >> I see that way? Reinvest, Reinvest. I haven't heard a customer yet that talks about reducing staff. Their I t staff is really, really critical. They want to invest up Kai throw buzzword out there, Dev. Ops. You didn't mention that it's all about Dev ops, right? And one thing that's interesting here is were or ah, technology that supports virtual environments and how many software developers use virtual environments to write, test and and basically developed programmes lots and being able to give those developers the ability to create new machines and be very agile in the way they do. Their test of is awesome and in terms of just taking big amounts of data from a nap, if I can circling APP, which is these virtual machines be ableto look at that on the infrastructure and more of her copy data so that I can do stuff with that data. All in the flying virtualization we think of Dev Ops is being very much a cloud thing. I'd say that virtual ization specifically server virtualization is the perfect foundation for Dav ops like functionality. And what we've been able to do is provide that user experience directly to those folks up the stacks of the infrastructure. Guy doesn't have to touch it. I wanted to pull >> a couple of threads together, and I think because we talked about the original vision kind of E m r centric, VM centric multiple hyper visors now multi cloud here in the world. So what >> are you seeing >> in the customers? Is that is it? Is it a multi cloud portfolio? What? What are you seeing your customers going to in the future with both on premise hybrid cloud public. So where does where does 10 tree fit into the storage portfolio? >> And they kind of >> fit all over the map. I think in terms of the most of the customers that we have ultimately have infrastructure on site and in their own control. We do have some that ultimately put those out in places that are quote unquote clouds, if you will, but they're not in the service. Vendor clouds actually have a couple folks, actually, that our cloud providers. So they're building their own clouds to service customers using market. What >> differentiates service is for serving better d our offerings because they can offer something that's very end end for that customer. And so there's more. They monetize it. Yeah, and I think those type of customers, like the more regional provider or more of a specialty service provider rather than the roll your own stuff, I'd say that Generally speaking, folks want tohave a level of abstraction as they go into new architecture's so multi cloud from a past life I wrote a lot about. This is this idea that I don't have to worry about which cloud I'm on to do what I'm doing. I want to be able to do it and then regards of which clouded on it just works. And so I think that our philosophy is how we can continue to move up the stack and provide not US access to our analytics because all that analytic stuff we do in machine learning is available via a P I We have ah v r o plug in and all that sort of stuff to be able allow that to happen. But when we're talking now about APS and how those APS work across multiple, you know, pieces of infrastructure, multiple V EMS, we can now develop build a composite view of what those analytics mean in a way that really now gives them new inside test. So how can I move it over here? Can I move over here? What's gonna happen if I move it over here over there? And I think that's the part that should at least delineate from your average garden variety infrastructure and what we like to call intelligent infrastructure stopping that can, Actually that's doing stuff to be able to give you that data because there's always a way you could do with the long way. Just nobody has time to do with the long way, huh? No. And I would actually say that you >> know what you just touched on, uh, going back to a fundamental 10 tree. Different churches, getting that level of abstraction, right is absolutely the key to what we do. We understand that workload. That virtual machine is the level of abstraction. It's the unit infrastructure within a virtual environment in terms of somebody who's running databases. Databases are the unit of infrastructure that they want to manage. So we line exactly to the fundamental building blocks that they're doing in those containers, certainly moving forward. It's certainly another piece we're looking. We've actually, uh I think for about three years now, we've been looking pretty hard of containers. We've been waiting to see where customers were at. Obviously Of'em were put. Put some things on the map this week in terms of that they were pretty excited about in terms of looking in terms of how we would support. >> Well, it certainly makes it more interesting if you're gonna lean into it with someone like Vienna where behind it. I mean, I still think there are some questions, but I actually like the strategy of because if I understand it correctly of Visa, the sphere admin is going to see the spear. But ah ah, developers going to see kubernetes. So >> yeah, that's kind of cool. And we just want to give people an experience, allows them to self service under the control of the I T department so that they can spend less time on infrastructure. Just the end of the I haven't met a developer that even likes infrastructure. They love to not have to deal with it at all. They only do it out. It assessed even database folks They love infrastructural because they had to think about it. They wanted to avoid the pitfalls of bad infrastructure infrastructures Code is yeah, way we believe in that >> question. Go to market. Uh, you preserve the 10 tree name so that says a lot. What's to go to market like? How are you guys structuring the >> organizational in terms of, ah, parent company perspective or a wholly owned subsidiary of DDN? So 10 tree by DDN our go to market model is channel centric in the sense that still a vast majority of people who procure I t infrastructure prefer to use an integrator or reseller some sort of thing. As far as that goes, what you'll see from us, probably more than you did historically, is more work with some of the folks in the ecosystem. Let's say in the data protection space, we see a rubric as an example, and I think you can talk to some of that scene where historically 10 Tree hadn't really done. It's much collaboration there, but I think now, given the overall stability of the segment and people knowing exactly where value could be added, we have a really cool joint story and you're talking about because your team does that. >> Yeah, so I would certainly say, you know, in terms of go to market Side, we've been very much channel lead. Actually, it's been very interesting to go through this with the channel folks. It's a There's also a couple other pieces I mentioned you mentioned some of the cloud provider. Some of those certainly crossed lines between whether they're MSP is whether they are resellers, especially as we go to our friends across the pond. Maybe that's the VM it'll Barcelona discussion, but some of those were all three, right? So there are customer their service providers there. Ah ah, channel partner if you want terms of a resellers. So, um, it's been pretty interesting from that perspective. I think the thing is a lot of opportunity interview that Certainly, uh, I would say where we're at in terms of, we're trying to very much. Uh, we understand customers have ecosystems. I mean, Marco Mitchem, the backup spaces, right? Uh, customers. We're doing new and different things in there, and they want us to fit into those pieces. Ah, and I'd certainly say in the world that we're in, we're not tryingto go solve and boil the ocean in terms of all the problems ourselves we're trying to figure out are the things that we can bring to the table that make it easier for them to integrate with us And maybe in some new and novel, right, >> So question So what's the number one customer problem that when you guys hear you say, that's our wheelhouse, we're gonna crush the competition. >> I'll let you go first, >> So I'd say, you know, if they have a virtualized environment, I mean, we belong there. Vermin. Actually, somebody said this bed is the best Earlier again. Today in the booze is like, you know, the person who doesn't have entries, a person who doesn't know about 10 tree. If they have a virtual environment, you know, the, uh I would say that this week's been pretty interesting. Lots of customer meetings. So it's been pretty, pretty awesome, getting a lot of things back. But I would say the things that they're asking us to solve our not impossible things. They're looking for evolution's. They're looking for things in terms of better insights in their environment, maybe deeper insights. One of the things we're looking to do with the tremendous amount of data we've got coming back, Um, got almost a million machines coming back to us in terms of auto support data every single night. About 2.3 trillion data points for the last three years, eh? So we're looking to make that data that we've gotten into meaningful consumable information for them. That's actionable. So again, again, what can we see in a virtual environment, not just 10 tree things in terms of storage of those kinds of things, but maybe what patches they have installed that might be affecting a network driver, which might affect the certain configuration and being able to expose and and give them some actionable ways to go take care of those problems. >> All right, we gotta go marry. I'll give you. The last word >> stated simply if you are using virtual, is a Shinto abstract infrastructure. As a wayto accelerate your operations, I run the M where, if you have ah 100 virtual machine, 150 virtual machines, you could really benefit from maybe choosing a different way to do that. Do infrastructure. I can't say the competition doesn't work. Of course, the products work. We just want hope wanted hope that folks could see that doing it differently may produce a different outcome. And different outcomes could be good. >> All right, Mario Graham, Thanks very much for coming to the cubes. Great. Thank you so much. All right. Thank you for watching John Troy a day Volante. We'll be back with our next guest right after this short break. You're watching the cube?

>> Narrator: Live from Las Vegas, it's theCUBE. Covering AWS reInvent 2017, presented by AWS, Intel, and our ecosystem of partners. >> Okay, welcome back everyone, this is theCUBE's exclusive coverage, live, in Las Vegas, 45,000 people here on the ground, for Amazon Web Services reInvent 2017. Their annual conference. Our fifth year doing it, I got two sets, two cubes, a lot of action. Day two of three days of wall to wall coverage. My next guest, Tom Kemp, CEO, of Centrify, security company out of California in Silicon Valley, leader in identity based security in the cloud, on-prem, big business growing, fast growing startup in the area. Good to see you. >> Yeah it's great to be here again. >> Security has been Amazon's kryptonite for many years. They've done their work, their paying their dues, they're checking the boxes. Certainly we see that on the federal side, public sector. Great success, Teresa Carlson, has done an amazing job. It's been fun watch her go from an outcast to, in the marketplace, "Ah, we don't trust the cloud", to winning. They've done the work. Security, you've gotta do the work. >> Yeah, I mean, they've done a great job of evangelizing the shared responsibiloty model where they clearly identify, "Hey, this is what we do", and then, "This is what the customer needs to do." So it's actually a very nice model that they offer that vendors such as us can slot into. >> And they move so fast but again, security is one of those things, you can't fake it til you make it. Right? (Tom laughs) You can't make it til you make it. Which means, it's hard. What are you guys doing with Amazon now? What's your story here for Centrify? >> Yeah, we're doing a couple of things. So the first thing is that we do privilege management. I mean the reality is is that the keys to the kingdom are in the AWS console in terms of the billing systems, firing up servers, shutting down servers et cetera. A lot of the more recent hacks have been because people have gotten the access to those keys of those systems as well. So we help lockdown the AWS environment and then we also help lockdown the actual servers being deployed on EC2. We provide multifactor authentication et cetera. The other thing that we do is and what we announced just the other day is we've actually moved our platform over to AWS. So before we ran on at Azure, can I say that at this, ah? >> John: That's fine. >> It's okay, yeah, just joking. >> All fair in love and sharing the cloud. >> So now we have a production cloud on AWS and we've also integrated in the marketplace. So there's SaaS billing that people can get as well, which actually is a very unique thing that AWS offers that the other cloud providers don't do. >> Alright, so I gotta ask you, obviously, to me, super exciting show because some of the announcements are really kind of cool and sexy, and some are under the hood geeky, like Lambda. And then you got the cool AI stuff happening, whether it's VR, AR, or recognition, all these cool machine learning, democratized toolkits. So does this help you? I mean Lambda server lists is a dream for a developer. Just, "Oh my God, I don't have to worry about anything. "What's a local host? "I don't need to know what a load balancer is." Does that help you guys or not? >> Yeah it does, I mean the reality is is that the amount of servers and applications, be it server or server-less, the amount of applications, the users that are connecting to it, it just adds more to the potential complexity. And we can, through the power of identity, provide a control plane to give people identity driven security and really allow people to move-- >> But it doesn't replace us. My point is, I guess, if you're locking down servers, this is a value right? >> Yeah. >> EC2 instances. But if the developers aren't using EC2 instances 'cause it's server-less. Are you guys transparent, are you abstracted away? >> So we also then, then integrate into the application and then help facilitate security for the actual users themselves. But look the reality of the situation is is that people are always gonna have a hybrid environment. They still have on-premises, which users have to access that environment. They're gonna have the cloud environment. And it's gonna be heterogenous. So AWS is a clear leader in the cloud but you're also gonna have Azure, Google, and then the SaaS applications as well, which are gonna be used in conjunction with the custom applications people are building. So the one constant-- >> I've been saying, I've been saying this for years, the specialty cloud is a big market. Oracle's a specialty cloud, Microsoft's a specialty cloud, 'cause they have apps for them. They can be different clouds. Multi-cloud is what's coming, would you agree? >> Yeah, and the reality is as companies go through digital transformation they're gonna open up more and more of their applications to more and more users. They're gonna be more and more devices, and that's just gonna lead to identity sprawl, more and more passwords that people have to deal with as well. And that's why in a world in which-- >> How bad is that problem? 'Cause that's a huge problem, at least in my mind. Identity sprawl, explain what that is and how bad is it? And what are the consequences if it's not fixed? >> Well look the reality is 80% of breaches nowadays involve compromised credentials. I mean we had the whole election, Podesta, the DNC, the recent hack of HBO, you had Sony. It always tied into people stealing credentials and people having too many credentials, sharing credentials, et cetera. So the problem that we face as consumers in terms of having too many user names and passwords has now entered into the actual enterprise and we're now in a situation that, yeah, there's an app for that but that means that there's a password for that. So IT is having a hard time controlling who can access what while end users are just dealing with too many user names and passwords as well. So you have identity sprawl, it's difficult to provision access. And then now you have IoT coming onboard and those devices need an identity unto themselves. And probably the thing that excites me most about some of today's announcements is what AWS is doing with IoT. Some pretty cool stuff. >> I mean I think IoT is the trend, AI and IoT, because, to me the data center, and this might be a little bit over the top, but I'll say it anyway. I think private cloud is real, the way Wikibon talks about it but it's still cloud and the cloud looks at these endpoints as edge devices. So a data center is just an IoT device, a big one. >> Yeah. >> Or, a series of devices connected to the network which connect to the cloud. I mean if it's operating as a cloud what's the difference? Private and public. >> Yeah, no, I, I, I-- >> IoT has gotta be connected. That's where identity could be helpful. >> Identity, I mean, 'cause look, every device has an identity beyond just an IP address. I mean some of the attacks have even taken over IoT devices and then pointed them against websites and brought those websites down as well. So users have multiple identities. Devices have identities unto themselves so you've got this kinda n-by-m, you know, situation where you multiply the number of users times the number of devices, and we're told digital transformation, more and more users are coming online connecting to applications. So I think that's a, it's just a great market to be in. >> Tom, great to have you on theCUBE, congratulations on your business growth. What's your secret sauce? We'll end this segment by you just taking a minute to describe to the folks watching why are you doing so good, what's your secret sauce, what are the tailwinds for you, why the success? >> Well the tailwinds are, first of all, identity has become the top attack vector. It's now involved, compromised credentials stolen at NEs is now involved in over 80% of all breaches. And the other tailwind is the whole move to the cloud that just says, introduces password sprawl. And we're very unique in the market in that we can secure both end users and their identities but we can also secure the privileged accounts that are built into the infrastructures of service. The AWS, EC2, IAM-- >> John: The critical resources. >> Yeah, and we do this in a hybrid environment. So, yes, people are aggressively moving to the cloud but you know and I know that still, what, 70, 80% of IT is still on-prem, and it's gonna be a mixed hybrid environment. And we offer both software and cloud services to secure both end users as well as privileged accounts in that environment. >> Alright, the bottom line, the AWS cloud phenomenon. Describe it in a sentence. >> In a sentence? Oh, it's just, the complete consolidation of all IT in a single platform. I mean, it's amazing that every year they announce another couple a hundred new brand new services as well. So it's just like a phenomena that I've never seen before in terms of a vendor aggressively able to come out with new capabilities and deliver more and more features. >> Cloud as an operating system that's what I always say. And I can see it coming together, and they're staying on their track. I gotta give Andy Jassy credit, even though I busted his chops by putting the Gartner slide on there, because that's old guard technically, doesn't match his presentation, so he's gotta fix that. They stay on their line, they're not wavering. They are mission focused. Changing the game, adding value for customers. >> And they're thinking about new app scenarios and I think it was brilliant that, take IoT, there's so many different flavors of operating systems for IoT. They're saying, "Hey, we're gonna come out "with a standard operating system "that you can leverage. "And we're gonna provide device management, "and we're gonna tie it back into the platform." So they're gonna capture the, they're trying to capture the edge. And the good news is stuff like that does provide opportunities for vendors such as Centrify. >> And they surround themselves with a great ecosystem. You guys are doing great in there. I know you're growing but you're soon to be bigger. But Intel, they're doing great with Intel. Intel gets a lift off this, more compute, everywhere. >> Absolutely. >> So even if they, they kind of have to split some of the business, whatever they do, who knows what happens there but Intel wins with this scenario. Amazon's not trying to eat the whole pie, they're sharing. They're sharing the wealth. And they do it, in the case of security again I go back to their shared responsibility model. It provides a great framework where it makes it very easy for vendors such as ourselves to say, "We play here, here, and here." So it makes it great to partner with and the ability for them to actually have SaaS based applications in their marketplace as well. And that's powerful, and no other of the cloud guys have a similar concept. Yeah, you could put AMIs on infrastructure as a service but to actually have a cloud based service tied into the billing system of AWS is incredibly powerful. We're very excited about being a part of that. >> And we will keep an eye on them on the open source side, certainly that's an area we're watching very carefully. Hey the developers love Amazon and that's a good thing. Now the enterprise love Amazon, public sector loves Amazon. Who doesn't love Amazon Web Services? We'll be following that very closely over the course of the next few months and next year, 2018. Of course live here in here in Las Vegas is AWS reInvent 2017. Back with more coverage after this short break. (upbeat electronic music)