 from New Orleans, Louisiana, at theCUBE, covering .NEXT Conference 2018, brought to you by Nutanix. Welcome back to theCUBE. Here in New Orleans, Louisiana, I'm Stu Miniman, we'd like co-host Keith Townsend who is the CTO advisor. And this is the CTO segment. Happy to welcome back to the program, we have Binigill and Rajiv Murani. Both of them are CTOs. Binni, you've got cloud services and Rajiv, you have cloud platforms. Let's start there. When we talk about, you know, there was a survey when you registered for the event and said, what do you think of Nutanix as? Am I your server vendor? Am I your HCI vendor? Am I your cloud vendor? Am I your mega Uber platform of everything? Got platforms and services. Help us understand a little bit how this fits and how you look at the portfolio and well, arm wrestle if you guys can't agree. Sounds good. You want to go ahead? So both of us obviously work very closely together but broadly speaking, I look after the core stack, the storage, networking, hypervisor, including Prism and then Binni looks more at the services you're building on top, the era, calm, things like that. So Binni, can you please tell us that a bit? Given the, you know, breadth of the ambition that we have, right? I mean, it's good to focus on the two layers separately in some sense, you know, build a platform that is capable of hosting a whole bunch of services. As you can see, you know, what Amazon and others have evolved, you know, they've spent a lot of time building a platform and if you think about it, even Nutanix for the last seven, eight years has done a really good job and once you have a solid foundation and building cloud requires some new capabilities as well. As Rajiv has said, networking security on top, now you can start building services and services themselves have a stack, right? Because there will be higher level services that use some lower level services in this. So that's a long journey ahead of us. Yeah, I mean, it's a great point because every time it seems like we have, you know, oh, this next generation thing, I'm not going to have to worry about the underlying thing. Virtualization is going to totally abstract it. We've spent a decade fixing the storage and networking challenges there. Containerizations, once again, it's like the application done there. Serverless, of course, will take care of all of this, but everything underneath it, it still needs to work. How do you balance and give us some of that? You know, what's the glue versus, you know, abstracting and going to developers? Maybe let's start with the platform. Well, the platform's always going to be there, right? And as we look at things like containers, that's actually where things get messy. How do containers work with storage is one of the bigger issues right now with the Kubernetes and other frameworks. So we have to start with the platform. We build on top of that and hopefully abstract enough that the services themselves don't have to deal with the messiness of the platform. Yeah, and if you look at how technology is evolving, you know, the more things change, the more they remain the same, right? I mean, the platform used to be Linux, Windows. I mean, that's the operating system on which I built my applications, right? Now the new platform is cloud, right? AWS is a platform, it's an OS and Azure is one OS. And how do you build applications that can run on these new next generation platforms? But the kind of problems to solve are still the same. I want to snapshot my application, back it up. I want to move my application one place to the other. I want to scale it out, scale it in. So the problems are identical to what we've had, right? But it's just that solving it with the new tools that we have, Kubernetes, containers and so on. Yeah, and sometimes birds just fly right through our studio. Yeah, I mean, we worry about bugs and now we have birds flying in too. So Rajiv, talk to us about, you basically have two different types of cloud customers. You have to serve many organizations. You also have to serve your external clients, doors, network compute, have to have APIs, have to have capabilities, basic capabilities that both your customers who want to build their own overlay and then Nutanix services on top. Talk to me about how do you make sure that you're building the best cloud platform to be consumed by cloud services, whether they're Nutanix cloud services or someone else's? I think just comes under the core principles that we have built the company around, right? That we will always build things around web scale design. So it has to scale to very large deployments. It has to be completely distributed. It has to go through a certain amount of wetting in terms of having APIs exposed. Nothing we do internally is through secret APIs. Everything is public APIs. So you're pretty stringent on some of these things. And then of course, layering on the simplicity of Nutanix is the other thing that we take very, very seriously. So when we do all that, nice patterns emerge. I think it lends itself to an elegance that the platform provides for the rest of the stack. So then we get to a confusing abstraction, which is, you mentioned earlier, containers. Who gets containers? Is that your organization? Is that your organization? Is it a fundamental part of the foundation, or is it a cloud service? I think the trick is to not necessarily worry too much about the boundary here, because frankly, this is something that the industry is still figuring out. What layer is this new Kubernetes thing at? Is it just that containers? But actually now it's going into all the way application provisioning, load balancing, and distributed routing, all sorts of things. I mean, we work as a team, essentially, and there's a whole bunch of engineers that are looking at the whole picture. It's always very important to look at the entire picture, and then figure out what are the right layers to go solve the problem. And when you're looking at containers, the bigger problem that our customers are talking about is how do you deal with the legacy plus the containers in one environment? Now I have my application, it's a three-tier application. The database, I still want to run in a VM, right? But I want to start tasting this Kubernetes thing, so I want to go with the web tier with containers, but it needs to be in one view, and that's what COM demonstrated. Through COM, you can orchestrate an application that's part VM, part containers with Kubernetes, and help our customers transition. So which layer these things are, and it's going to be an evolving answer. So, Vinny, I love that you started the conversation around COM. Is COM the first interaction that most customers will experience when it comes to Nutanix cloud services, or is there a different, one of the other services, the more likely first experience of cloud services versus the traditional compute storage network? Right, so the first cloud service that we have announced that we'll deliver is DR, right? I mean, that's the first one with Xi. Once DR is available, very quickly we'll add more services. Beam is another one that has to fold in into Xi cloud services. When I say fold in, it essentially means you have the same identity, and you have the same billing mechanisms, and the same experience. You know, similar to when you go to a public cloud, you'll see there's a host of services, and they're sort of equals, and you can pick whichever one you want to use. What we want to provide with Xi cloud services is that the same experience, except that these services are now hybrid. You can have them on-prem, you can have it in the cloud, and our teams are building this hybrid view. Some of it, the preview of it, what you already saw in the demo where you saw available to be zones on both sides shown on one screen. Now you will see the service footprint on both sides on one screen. Yeah, return experience point of view. I think COM will be how people will see this for the first time, because that's going to be the center marketplace that we will have. So people will learn services from this. Right, so where's the portal for cloud services? Right. As I understand it, COM is that portal. COM is a lot more than that. It will have not just services, but applications and workloads as well. But yes, the experience will start with COM. So when we talk about a hybrid cloud world and the platform, people are trying to understand what exactly lives where. When we hear kind of Xi, I wonder if you might be able to give us kind of a compare contrast of, say, that you look at VMware and VMware and Amazon is kind of an easy one to understand as it's relatively the same status living in a different data center. Right. So we're doing things a little bit differently. While we're building our own cloud data centers today, we are architecting in a way that we're not tying it down to any single stack that it has to be only a Nutanix oriented stack. We absolutely intend to scale this out by partnering with service providers, with cloud vendors, and so on. We saw something in the keynote yesterday about running nested on GCP. You can imagine where that will go in the future, but other clouds also on the radar. So much like we did with our HCI stack, we shipped on Supermicro, but very conscious of the fact that it's software that we can move anywhere. We have buildings that are exactly the same. And what I'd add is while we are doing it in our own data centers right now, we are learning a lot, right? And as we are learning the things that are truly needed to make running a cloud easy from an operational perspective, that allows us to build a product that is an honest product to give to our partners and service providers to say, now you go run it and you won't be spending too much. Like for example, the experience that they've had with OpenStack, it cannot be repeated again, right? So that's what we want to do. So let's talk about the relationship with Google as a model going forward. Is that prototypical what you're looking to do with other public cloud providers? And first, give us some color around the announcement. We have a heavy one on the cube. Talk about XI in Google and then kind of the strategy moving forward. Yeah, so a lot of the public cloud vendors are actually realizing that hybrid cloud is important, right? And as part of that, they're providing bare metal services and Google has this nested service to enable others to bring their own stack, you know, virtualization stack to run there. Amazon has done it VMware. Amazon has also announced their intention to GA bare metal services. So we see a future where a lot of these public cloud vendors will offer bare metal and that's where our XI stack will run, right? And also giving customers choice to go from one cloud to the other seamlessly. Today, we know that Nutanix can move from public XI cloud to on-prem and back, but once you have XI cloud running on multiple cloud vendors and you can move between cloud vendors seamlessly as well. And that's a really compelling message for our customers. Great. One of the challenges for some of us watching is you've got a pretty big portfolio now and some of the things out in the future and it's like, okay, where does Nutanix fit? How do they have the right to participate in this? Wonder if you can talk a little bit about ERA and maybe Sherlock is a little bit further out. So ERA is about managing copies of your databases, right? And again, if you look at where a lot of cost is sunk in enterprises, I'm running my database, a production database. For every single production database, there'll be maybe tens of test copies of it. What ERA does is minimizes the cost of managing the copies and also it's thinly provisioned copies. That's something that our customers have said that that's a real pain point for them, right? That nobody solves really well. So we decided to work on that. That's just a starting point of what we can do in this PAS layer and also helps us learn this space as well. We are reaching out to not the infrastructure admin but actually to the database admin, gives us a new audience to talk to as well. So from an audience perspective, we are broadening the scope. We are reaching closer to the lines of businesses and the decision makers, which is good. Now going to Sherlock. Actually, if I could just one quick follow up on the database piece. Database migration is really hard. Talk to any customer and you say database migration and it's one of the things that strikes fear in them. Talk just for a second if you could about the expertise that your team has and why you believe you can really deliver that push-button simplicity that Nutanix is known for. Oh, so yeah, the team that's building ERA are hardcore oracle folks who have decades of experience doing those kind of hard problems. And they've come here with a mission into Nutanix that we are going to solve it. Using the Nutanix platform that we have built, there are so many things that can be done in a better way. And since we have a clean slate, we can start afresh and do it right way. So yeah, from our capability to do it in the right way, making it simple for our customers, we don't have a doubt in fact, a lot of customers who have tested this in alpha, they have raving reviews on that and they just want it as soon as possible. And on the database migration subject, we also have a tool called SQL Extract that we've been shipping for some time that helps you migrate your databases from existing 3D or even hyper-converged stacks onto Nutanix and so we have some expertise in the area already. So a little bit on it, I heard in term copy data management. Is this mainly copy data management or is this actually database migration to ability to move from one database to another one? Or is it all of the above? So it's doing management of copies. It's also allowing you to clone databases, right? So you can go to a snapshot and clone another one. Migration is not yet there, but it's a natural consequence of the capabilities that we have because once you have snapshots, we have the capability of moving snapshots from one data center to the other using our DR capabilities. So that's on the roadmap. Then further down the roadmap is database provisioning itself. If you want to provision a brand new database, you can also do that. So these are just natural transitions of what, but what we wanted to do, just like what we did with Zai, start with the hardest, torneous problem and then work backwards and do the simple things. All right, so unfortunately we're running short on time. Give us a closing word. I want Rajiv and Vinny. Maybe you could talk a quick second about Project Sherlock and give us some things that we should look forward down the road from Nutanix. Yeah, so we believe that the world needs an enterprise cloud operating system. What it means is it can run on my private cloud, in the public cloud and on the edge and Sherlock comes there. I mean, it's taking our stack and creating a mini pass version, as you saw in the demo, and running it at the edge in a way that all of your footprint appears like one dispersed cloud. And that's a pretty exciting space and we think that is the key differentiator that we'll have going forward. All right. And any final words for Rajiv? I think we covered a fair amount of ground, so yeah, thanks for having us on. All right, well, it goes back to really that distributed architecture that is at the core. Appreciate having the conversation. The CTO roundtable, as it were, Vinny Rajiv, always a pleasure to catch up. For Keith Townsend, I'm Stu Miniman. Back with more here. Thanks for watching theCUBE.