 Francisco, extracting the signal from the noise. It's theCUBE, covering VMworld 2015. Brought to you by VMworld and its ecosystem sponsors. Now your host, Stu Miniman and Brian Gracely. And we're back. Here with SiliconANGLE TVs live continuous wall-to-wall coverage from the Moscone North here in San Francisco for VMworld 2015. I'm Stu Miniman and joined with me. My co-host on the director set is Brian Gracely. And we're talking about the intersection of cloud virtualization, networking and storage. Excited to have with us, Benny Schneider, co-founder, chairman and president of Revello. Thank you for joining us. First time on theCUBE. And Dave, nice to be here. All right, so Benny, that's a lot of topics. We talk VMworld is really, you know, that the ecosystem for virtualization touches on a lot of environments. A lot of storage people here, a lot of people looking at how cloud is transforming what they're doing, trying to sort out what this whole hybrid cloud thing means. And Revello has an interesting place in kind of the hybrid cloud discussion. So maybe bring up for our audience that aren't familiar with Revello. You know, what your role is in that ecosystem. Okay, so we are coming from a very rich virtualization background, networking background and some storage. And in our previous company, KumaNet, we did KVM. It's not keyboard, video mouse. It's kernel-based virtual machine. It's the famous hypervisor that many of the new upstarts that you see here in the Moscone Center are using. Google is using it based on the experience that we have in networking and the success that we have with KVM. We have built a new vision on how the future data center should look like. And the new vision is basically based on isolation of the application and the infrastructure that the application is running on top. And when I say the application itself, it's everything that's related to running this application. It's the VMs themselves. It's the networking. It's the storage. Some dynamic rules and some operational aspects on how the application should behave dynamically when it's running. So for us, the hybrid is a very natural use case for our vision because we are not talking about migration. Migration is a singular event where you're running it one time here and then you're running it on another place. Our vision is that you will be able to run your application anywhere you want. You can run it today here. Tomorrow you can run it there. And the way to do it is in a very elegant and very sophisticated way based on our nested virtualization technology. Basically, you can think about it. We have a compiler that takes and map the application itself into the target cloud. So all this magic is done without changing the application. It's done automatically by Ravello with some key knobs that you as a user can play with and adjust your needs. Yeah, I tell you, Benny, in many ways, I've looked at it that infrastructure is held back the way that we handle our applications. One of the first times I talked to Docker, Ben, the CEO said to me, what we're going to do is we're going to separate the application management from my infrastructure management. I listened to the keynote this morning from VMware and they said their unified hybrid cloud message is their cloud with any app anywhere. So the high-level messaging from some of these pretty big companies, that sounds a lot like what they're saying. So can you talk a little? You said it's not a migration tool, but what are you doing? How are you doing? And how do you interrelate with the VMware and the cloud guys out there? Right, so I think we have a complementary solution. VMware is focusing right now the hybrid solution on the private cloud and VCA VCloud Air. We can run on VCloud Air. There is no reason that we cannot. We are right now focusing on the public clouds, mainly Google and Amazon. And we have the ability basically to take any workload and run it, as I explained before, with the new concept that we have created. We are doing it with the nested virtualization technology. So if you wanna think about the vision and how it's being implemented, let's compare what we are doing with nested virtualization to what we have done in the past and what VMware has done in the past with virtualization itself. So virtualization is basically a layer that's sitting between the VM and the infrastructure itself. Virtualization is doing three things to the operating system. It's detaching the operating system from the hardware. It's making the operating system the VM monolithic, which means in one file you have the binary, the executable, the configuration and the data. Once you have the two things, the separation and the monolithic notion of the application, you can move the application from one physical machine to another physical machine. What we are doing in Rovello is exactly the same thing one level up. Rather than dealing with a single VM, we are dealing with an application. Rather than dealing with a physical machine, we are dealing with the cloud. So we are making the application monolithic. We are detaching it from the cloud and then we can move it around from one cloud to another cloud. We are cloud independent, so to speak. The other thing that's unique about the Rovello vision, we deliver our technology as a service. So you can look at us like a virtual cloud provider, you know, like over the top cloud provider. We are running over the top of existing cloud providers. And you come to Rovello, you point to the VMs that you wanna run. There is a tool that compiles the overlay network for you, the overlay storage for you. You can of course go and edit it. Once you're ready, you can deploy it based on parameters that you provide or based on optimization that we can choose for you. So you have a way either to choose performance optimized, in which case you will choose where you wanna run it. There will be of course a higher price for it. If you choose to run it in a cost optimized, we'll choose for you the most cost effective infrastructure to run on top of it. Now we, so first off, when we had sent out some messages out to our community, we said, who do you guys wanna hear from? Who do you, and overwhelmingly, especially the Vexpert community, which is really sort of more the top-level technical people, they really dig into it. They said, we wanna hear from Rovello, we want you guys to talk to Rovello. As we talk about hybrid cloud, and you talked about, you know, you're sort of targeting the public cloud, it's interesting to me that more and more people aren't saying, aren't realizing this. Our research is showing Amazon and Azure and Google to a certain extent are far away from anybody else in terms of what there's there. How much do you hear from your customers that you have to be able to work with, with these other clouds, whether they're AWS, or OpenStack, or anything else, as opposed to just this, you know, hypervisor to hypervisor makes a hybrid cloud? We are not, to be honest with you, we are not hearing direct requests to run the workload on some specific clouds. We just hear from them that they want to have some solution to the problem that they are facing. And they, some of the problems are to run it in certain locality. So if they are running, let's say, training in Australia, they want the data to rule and the old workload to run in Sydney because they want the user experience to be local. We hear performance issues, we hear some regulation that they care about. All these things play together in part of the cues that they are providing to the optimizer that we have, we call it cloud map, that basically maps the application into the cloud itself. The VExpert community has been great to us, both in terms of being early adopters of our technology and in terms of providing guidance to our roadmap. You know, always when you have such set of engineers like we do, coming with a product or with the technology, it's relatively easy, it's all about, how do you make a technology and a solution that people can use, need it and can consume it? And that's what VExpert are helping us in achieving. So right now we are focusing with the VExpert on some specific use cases. These are virtual labs, virtual training. So rather than keeping their own infrastructure on site or in the basement or having some favors that people have to bring to them, they're using Revello. And Revello, we have a program for 1,000 hours per month that people can use it for free if you belong to the VExpert community. And if not, it's relatively cost effective solution relative to the alternatives. What we have done with the VExpert community, we have graduated from the beta program that we have and now we are going into the next step, we call it infinity data center. So I would like to explain what it is that we are doing with the new program. So first of all, the one that we graduated from and we announced GA just for this show is the ability to run the ESX hypervisor on the public cloud. And some of the cool things that people have done the VExpert community were when we got it just out of the box like three, four months ago, they were able to take and do the motion, life migration from the private data center to Amazon public cloud, then to Google public cloud and back to Amazon public cloud. Kind of a real hybrid VMotion solution. Yeah, so Benny, I'm wondering if you can unpack that a little for us because those of us that have been using looking at cloud for a while, there was always this discussion of like, oh, we're going to do cloud bursting. Cloud bursting really doesn't work. It's not easy if you have something with storage. I can't move it fast. It's not follow the sun. Right. Many of these things, can I spin up instances in various places? Sure. Can I have an application? Sure. Do I maybe need to move some of the compute to the storage? Absolutely. Analytics is a great use case for that. So I think I understand you want to have flexibility and sometimes you will need to move, but we're saying it's not migration. So can you help us kind of understand the use case and where and why they're doing that? That's a great question. There are two ways to look at it. One way is like we're discussing so far and that's the ability to run the application as a monolithic entity. You can run it one time here. You can run it one time there. We have customers that are actually doing that. One good example is a checkpoint that have done training for 1000 SES. And each one of them got is on, if you want capsule or is on application, private one, it was really bursting because you couldn't run it on any public cloud, even the big guys. So we started from one location. We moved it to another location. Once we exhausted the resources on each one of this cloud, we moved it to another cloud. So this was a typical use case where the bursting to the cloud using the cloud economics is based on the application level. What we are talking about now is a different view or different way to look at the bursting. So think about in the past, when you were running out of capacity, you would typically buy a server from Dell or from HP. You will install ESX on it and you will use vCenter or vSphere to manage those resources. With a new product that we are coming in beta right now, we have the ability basically to allow customer to rent the hardware from Rovello. So you are renting the server. You are using the cloud economics. You are installing ESX on it and then you can manage it from the same glass. So from VMware perspective, from the management system of VMware, you are still managing VMware resources. The only difference is running on a virtual hardware, virtual silver, rather than it's running on a physical silver in your data center. So to me, this is one great way to do bursting because let's say that you have now a need to run some over the weekend test and dev for a release that's coming next week. You don't have servers. You don't have IT resources to run them. You cannot wait for the hardware to be shipped and install it. You're coming to Rovello. You are renting the hardware. We are installing it to you and nothing changes. You are managing it the same way. Same methodologies that you use for your agile development continue to serve you exactly the same way. And in this case, we are actually walking together with IT guys to be able to use the very same tools, methodologies that they used before. So in essence, you offer people the opportunity that says, look, if I look at a typical application lifecycle, development, QA, product, test, staging, a lot of those are sort of temporary in some cases. Now you may update them, but other than production, that's a long term type of thing. You're basically telling people, let me rent a lot of that workflow that for a period of time, you can buy the production. And then for things like DR, I'm assuming that really is probably more economical to be a rental model as well. Is that the way to think about where Rovello fits into somebody's application lifecycle? Absolutely. So I would view it in the following way. The cloud makes sense economically in certain use cases. For example, I'm coming from Israel when I'm lending in SFO, I'm renting a car. I'm not buying a car, even though for a day it will be cheaper for me to buy a car, but I'm not going to use it for the whole year. So if I have the same logic and the cloud economics works for me, I want to leverage it, then I would better off rent a car or rent a machine if the workload is bursty, like the ones that you mentioned, the DR, the entire lifecycle. I can see this going also into production use cases. It's not only for non-production. I can think about bursty workloads. We have some gaming companies which have tournaments that are running on a periodic basis. There are some broadcast events that make a lot of sense to run them, even in production, using the cloud economics. This is where we are going. So if you look at the overall roadmap where we started, we started from running VMs in the cloud. Then we went and moved the entire thing, the entire application, everything that's related to your application, the way that we are looking at the application. Now we are doing it with the hypervisor, so we are doing less than two. We can run the entire hypervisor. One good example is OpenStack. We can run private cloud OpenStack with all the KVM or Zen hypervisors on top of any public cloud. Reddit is using this extensively for training. And we are moving now into the Infinity Data Center, which is focusing more and more in production use cases. All right, so, Benning, I mentioned Docker before. I'm curious how containers fit in, even to talk about LXD coming out, giving us some kind of the V-motion-like capabilities in container. I'm just curious how the container mindset and technologies fit into what Ravel is doing. So I think it's a complementary technologies. I don't think that they are competing technologies. There are cases where containers make sense, and there are cases where virtualization makes sense. Typical use cases today for application that are not so-called born in the cloud are a variety of operating systems that cannot run on container as is. They cannot run on the same host. There are some issues related to security and so forth. So we are today to approach it, even that VMware is proposing to deal with container. One is native architecture, and the other one is just enough OS where the containers are running within a VM. In our case, we are agnostic. You can use Ravel to run both of those architectures. I think the part that people are missing on containers and virtualization is that virtualization is really the best way for IT guys to operate their data centers. It's not so much for the developer and the devop guys. It's more about the IT guys of having the infrastructure uniform. So all the components that the virtualization is managing are the same elements. These are virtual machines. It can be any virtual machine. It's much easier to a IT guy, IT shop, to manage infrastructure from the virtualization perspective where everything is the same rather than having an host for Windows and host for BSD or another host. Now, that being said, we are using containers in Ravel as well, and we actually have a really good way to combine between the two concepts together. Yeah, all right. So, Benny, I guess the last question I want to have for you is, the company's been out for about four years now. Yeah, a little bit over four years, yes. You know, what would you say to people that have seen Revello, as I said, when initially I ran across it, it was mostly for migration type technologies. What do you want people coming away from VMworld to understand about Revello technology and the company as a whole? I think the important thing is that they have to think about infrastructure, independent application life cycle. So you develop your application once. If you're comfortable with containers and you can afford to do containers only, go ahead and do containers. If you want to do it with virtualization, go ahead and do it with virtualization because you have different VMs. The other part is that we don't require you to rewrite your application. You just take the VM as is and you can run it without any modification. You don't have to rewrite your application. It's not that we are forcing you to write in a certain way. The other thing is we allow you a good way to consume the IT in a way that's the cloud intended to. Because the cloud at the end of the day is about cost saving. Not all the application today can use the cloud as is because they have different assumptions when they were created. We allow you, without any modification, to use the application as is. The last thing I want to say here is that when I started my career, I did some chip design and I did some compilation work. And to me, it reminds me of the old days of people writing in assembly. This is really old. And then compilers came on and people said, look, there is a better way to find somebody else to take care of matching your code to the underlying hardware. And by the way, if you use compiler, then you can match it to any architecture. That's actually our vision. We are the compiler of taking application, new, old application, and making the best fit into any infrastructure, private, public, any public. All right, well, Benny Schneider, thank you so much for joining us for this segment. Thank you, a great example of working with the Vexpert community. Thank you very much. And thank you for all your support. All right, thank you so much. So we'll be right back with lots more coverage here from VMworld 2015 for Brian Graceley and Stu Miniman. Thanks for watching.