 Live from Las Vegas, it's theCUBE, covering Discover 2016 Las Vegas, brought to you by Hewlett Packard Enterprise. Now, here's your host, Dave Vellante. Welcome back to Vegas, everybody. This is day three of our coverage of HPE Discover 2016. This is our sixth year covering HP, now HPE. We've been documenting the evolution, the transition, the acceleration now of HP's business. Vish Malshan is back. He runs product management for HP Storage and Mohit Batnagar is here. He's the vice president of products at Cluster HQ. Gentlemen, welcome to theCUBE. Thanks, Dave. So tell us about Cluster HQ, relatively new company. He's got some financing earlier this year on top of some earlier financing. So you guys are hitting your stride now. What's going on with the company? Yeah, so we are a startup. We are very passionate about delivering a whole vision of containers to enterprises. As we know that containers have been the sole of what, how Google delivers their technology. What most people don't realize that every time we are using our Gmail or doing some kind of search, Google is firing up containers. In fact, one of the releases came two billion containers per week. But what we are saying is that same vision can be brought into enterprises. And that's what we are passionate about. Cluster HQ, we are funded, well-funded. We are building, I think, one of the world's best team, which understands the old stack. You know, the VMware, HP, and the old thing, as well as the new stack. The DevOps, microservices, public cloud, hybrid cloud infrastructure. So that's what we are doing. And our focus is on data management part of it. The number of problems in containers that needs to be addressed. And we are saying, you know what, let's solve the data management problem. Okay, and then Vishu, you and I have talked about containers. Of course, containers have been around for a long time. Anybody who studied Linux, the containers were there, and all of a sudden there's resurgence. So what's your angle? Why does 3PAR, HP Storage Care? So Dave, I think you're seeing an emergence of what we call this new application paradigm, right? And we're hearing that as customers develop these new applications, they're looking for hyperscale, they're looking for cost advantages, they're looking for latency and performance advantages. And so this is a new method for developing those applications. Yet we also hear that persistence of storage is a key element, especially as you develop containers in one environment and want to maybe deploy it in a secure, encrypted environment, or you may want to deploy it on our flash environment. And so it's more indicated, right? I mean, like we had very close integration when we had the era of server virtualization, this new era of containers in the x86 space, I think brings open another new opportunity, new frontier for us to go look at. So Mohit, talk about the problem you guys are solving because of course we understood well the server issue and virtualization helped solve that and we sort of tried to push that into applications. Didn't, wasn't really the perfect fit, but talk about the problem that you guys are solving. Yeah, so let's talk about real quickly why containers because if you understand that, then I think we can talk about the problems that we need to solve. So most of us live, it started our life at one point or other, we deployed things which were bare metal. VM technology allows us to get much higher efficiency, et cetera. Continuous in that sense takes it one step further. Essentially what's happening unlike a virtual machine which requires a host operating system, it's kind of a lot of lips, et cetera, get loaded. Containers has shared operating system, so it becomes much lighter weight. But what does it mean? It means that you have now much higher density, so you can get container density which could be 10 times higher than that of a virtual machine density. Secondly, because the lightweight, you can spin up very fast, which means when you want to spin up a large number of compute resources, you can spin up containers. Well, one of the problems that happened with containers were if you have workloads which are stateless, where something is done and a call is made outside the container and the state is not critical, you know what container dies, it dies. Well, we looked at it and we say, wait a second, you're looking at solving a global problem, you want to go and build a unified cloud infrastructure. Why don't we actually enable state persistence into containers? And that's where cluster HQ came. In DockerCon last year, we made a contribution with a concept of volume plug-in came. And since then, we have done a number of things where, for example, using HP three-par systems, we can take all the goodness of HP three-par, high performance, all flash, reliability, disaster recovery, stability, SSD level encryption. And on top of that, we can enable any kind of container-based workload, which means whether it is Oracle SQL or the new generation database services, such as MongoDB or Radeus or Elasticsearch or Cassandra, across the board, we can deploy these stateful containers. The beauty of this is that organizations now can make an investment, which is whether it is a bare metal deployments or container deployments or virtual machine deployments, and using technology, which is a combination of our product Flocker, which is a volume manager, which works with HPE's product. Organizations can adopt the container-based workloads. So how did the relationship come about this partnership, Vish? I mean, we see Docker on stage this week, several years after Docker comes to the prom, and it's okay, that makes sense, right? They've proven themselves, and now, okay, we're going to put them to the HPE Distribution channel. This is different. You guys are early on in the game. Why? What did you see? I'll tell you, Flocker has announced the availability of joint solutions actually six months ago, right? And through the marriage of OpenStack, Cinder Drivers, and Flocker, we were able to bring that sort of class of persistence to the environment. So I think we found a common intersection. Moe was indicating to me that as these customers were developing, they got stuck in trying to get into production, especially as they're moving into IT, and you have three power on the floor, you have store virtual on the floor, for example. And so we found that was an intersection point here that we could participate together. So Moe, why three power? Why did you guys focus on three power? So, look, this space is going to have heterogeneity. So it's inevitable. And I think there are going to be other systems. People might use AWS or something. But when we look at it, what are the problems enterprise customers are going to solve? Containers is an interesting part of the equation. So obviously from a DevOps, microservices architecture point of view, almost every Fortune 1000 company is going to adopt containers. But that doesn't mitigate all the other requirements they have. They still need, as I said, HA disaster recovery, encryption, reliability, backup, and all of those things. So we looked at it and we say, who are the companies that actually understands that? So that was one part of it. The second part of it that enterprise, as Meg talked about, is that they're going to solve these problems across the stack. So the hybrid cloud is a reality, right? So we said, which are the organization which actually are going to deliver and solution that will solve those problems. So for us, three part became a very good choice. And frankly, I'm a storage guy myself. So when I look at it and I say, which is a system that has a range of capabilities. Yeah, there are other choices, but I think three part is a fantastic track record. And I think it is something which could become a dominant technology in this space. So that was our motivation. Well, storage business is changing pretty dramatically. I mean, you guys have both been in the business for a while. It used to be, okay, I buy EMC for block and I buy NetApp for file. Next question, right? That's changed pretty significantly, hasn't it? I mean, you're seeing the emergence of Flash clearly as one that's still pretty much in the system-defined space. You're seeing the emergence of software-defined storage as well, right? And then you're also seeing the marriage of both, right? Because I don't think it's in either or. And I think there are spaces where customers are going to say, look, I have mission-critical mainstream applications that want to develop one way. I have cloud-native applications that want to develop another way and deploy another way perhaps, right? And then you're going to see maybe some bridging between the two, right? As they deploy and they have specific requirements. And as Mohit pointed out, right? If you look at the three legs that we always talked about, Dave, the performance, affordability, enterprise-class data services, you know, I view this flocker, docker integration as part of that enterprise data services, right? Because there's a big intersection with reliability, whether it's replication, priority optimization. If you want to apply QoS-type rules, those are the kinds of things that they shouldn't have to go reinvent again, right? And if you draw a parallel, VMware, when they did their replications, they built this model with side recovery manager and side recovery adapters, right? And I think they very intelligently leveraged the ecosystem for replication. So I think it's a similar model here, right? And there's a Gestalt going on. I mean, Benioff got it right when he said that there can be more SaaS companies outside of the technology than there are inside of the technology business. And so everybody's becoming a software company. It's all about building platforms, APIs, agility, so the DevOps thing is taking off. Spot on. Spot on. So let me give you a real-life example. We have a customer. These guys, about a year and a half back, used to do application development, deploy maybe once in four months, model of the application. You'll say that's pretty good. Well, it started going to microservices architecture, broke down the entire model of the application, started doing deployment once a week, 10 deployments per week, currently 40, 45 deployments per week with a goal of 100 deployments per week. Now, this is a company. This is not a Fortune 10 company which had through all kinds of resources. This is a startup. Now, what this is giving them is that their DevOps guys are able to go, their developers are able to come up with ideas, implement fast. Now, if I'm to talk to the CIO, that company, he want to abstract away the entire infrastructure from the developers guys. And that's where the fact that they can take a three-part kind of system and say I give the enterprise class reliability. I deploy a container-centric workload which gives me the fast performance and density. And by the way, if I do an application development which is in public cloud and then eventually I need to bring it to on-premise, I can also do that. So I think these flexibility are, a lot of these are driven by forces which are application development accelerated cycle, the need for cloud-based instances. And that's where I think the combination of what new stack like ours is doing with the capabilities like three-part becomes very attractive. Well, the other major thing about developer productivity that I'd love to get your guys' thoughts on is data access and data sharing. Being able to, let's say through flash, be able to share copies of data that are current instead of so N minus one, we've all been through it. You're looking at a development project, oh jeez, I wish I could see that with some more current data. If similar, okay. How is that affecting and changing the developer mindset and the productivity? One answer. Yeah, sure, let me start there first and then I know more, you've got some perspectives there as well. So clearly, making sure that you can replicate as close as possible to the production environment with production type data is, I think it's a big requirement for development, right? I think that often struggle with, they can really never replicate the production environment or it's not a cost effective to replicate the production environment. And so, flash changes some of that because now, again, you have enough performance headroom, you have software capabilities to make copies effectively and then you can put priority controls around that to ensure that if I'm developing on a production system that I don't ever impact the production performance level, right? So God will. So I think, married now with maybe some of the concept that Moira is going to talk about with data, I think that's an interesting intersection here with DevOps, right? So we are super excited about the fact that while data has gravity, data can also have a temporal aspect to it. So the notion of snapshots in the past was a linear snapshots were taken. They were typically taken for disaster recovery kind of reason. Well, imagine, just like you have a GitHub for code, what do you have Git for data? What if along with the code, I can capture the state of data? I can give you a full graph of it. People can go, they can look at a particular snapshot. They can clone it. They can create a branch on it. They can merge that branch into something else. They can, if you can do that, suddenly it creates an architecture where microservices are no longer doing development in isolation of each other. They can almost have a Git for data, which is a common repository of data where things can be pushed and pulled. So that's a concept which we feel could be very powerful vision and we are working towards it. The beauty is what you said earlier because of efficiency of snapshotting, because of flash that it does not have the impact on it. You can actually truly create these copies and then you combine this temporal aspect of snapshotting with the spatial aspect of snapshot. I can move them across the clouds. If I can do that, suddenly we have taken care of the data movement in a very efficient manner as well, across time as well as space. I love this Git for data sort of metaphor because then there's a monetization model behind it where you're selling tooling on top of that. Right, absolutely. Beautiful. All right, we have to leave it there. I'll give you guys each a sort of last word. What's your takeaways from Discover 2016? Look, I've been in the industry 20 years. This is one of the best shows. I, congratulations to HPE team. I think they've done it. I can feel energy around it. What I also like is that instead of just saying, here is a product, here is a product, HPE is stepping up and saying, look, we solved your problem in a holistic manner. So we are super excited about this partnership. My message to our customers and HPE's customers is that the bold containers is a new territory, but those of us who have been in the industry for a long time, we take this stuff seriously. We are delivering enterprise-class solutions. So come along with us and we'll be with you during the journey. Great, and Vish, you came over from 3par. Very exciting company. When I got back into the business a while back, somebody said, look at 3par, they're the next big thing. This was quite some time ago. Brought that to HPE, transformed that architecture to an all-flash array without having to buy another company. That's got to feel really good. So tremendous success, growth, you're now number one, at least for a little while, until that merger goes through. Unless you're going to buy NetApp, I joked, but that's not going to happen. Just kidding, okay, it won't happen. Don't write that. So give us, give us a feeling good, spring in the step, give us the drop the mic quote, what's happening as you exit HPE Discover 16? So Dave, I think staying focused on customers, staying focused on innovation, and making sure that we can deliver a value proposition, making a difference. That's what we stand for, right? And I think we want to continue to ensure that we can make that difference for our customers. All right, well theCUBE, hopefully making a difference in bringing you all the signal at HPE Discover 2016. Keep right there, everybody will be back with our next guest. Thanks, you guys appreciate it. Thank you very much. Good.