(upbeat techno music) >> Welcome to The Cube. It's a wonderful Tuesday and we're here talking to Craig Nunes who's the VP of marketing at Datrium. >> Good to be here. >> And Craig, you guys had an announcement today and the announcement particularly refers to the further convergence or the opportunity to further converge not only hardware but now increasingly operating environments specifically bringing some of the Red Hat ecosystem over to the Datrium product set. So why don't you tell us what happened? >> Sure, we've been making a great business with customers in the VMware environment. We debuted our new generation of convergence back last year and as we were picking up customers in vSphere, we're running into a number of them who were saying, "You know, God, this is awesome. I do have "some Linux stuff going on. "Can you guys help me out there? "I can't seem to find a modern converged platform to really take on both environments." And so that's precisely what we've done. We are announcing today that we've partnered with Red Hat to use their stack, Red Hat Enterprise Linux, and their full Red Hat virtualization stack, run that on our DVX on our compute nodes alongside vSphere servers. Beyond that, because we observed there is a lot of activity going on in the container space. >> Peter: Just a little bit. >> CICD is becoming something that more and more folks are moving to. We've also partnered up with Docker and we're also going to provide bare metal container support with persistent volume plug-in for the platform. So this is all in one go, you now have really for the first time, a modern converged system that can handle what you're doing today with vSphere, probably handle what you're already involved in, but you're looking for way to bring this stuff together in your Red Hat environment. But then more importantly, you're kind of set up for where you're going with containers. >> So, when you say handle, Datrium has made some interesting decisions regarding how to solve some of the engineering problems associated with convergence. >> Craig: Yeah, yeah. >> Take us through a little bit about what it means to handle. >> Craig: Sure. >> What were you doing on VMware that you're now especially doing on the Red Hat ecosystem and will be doing as you move more closely towards containers? >> In the world of converged infrastructure, of course we started with kind of packaging convergence with arrays and servers. Hyper-convergence came along, really bringing storage into the x86 architecture, super cool idea in principle. The challenge with that is because storage is now part of your server, everything is stateful. Everything is a storage node and it's tougher to scale, it's tougher to service. Taking nothing away from the hyper-converged guys, it's great for single use case, great for edge, but we're really aiming for what people are trying to get done in the private cloud data center. So for that, we found that by separating the persistence, the durable capacity from the IO processing on the server, we could provide this wonderful converged platform that scales, that you can use any server you like, you can bring your blades, you could use our own compute nodes, whatever. It gives folks just a lot more freedom to get the job done. Servers are stateless like they were with your arrays but have all the benefits you're desiring with converged infrastructure. So, we brought that to vSphere and what folks have taken away is, "Wow, since everything "runs local on the server and Flash, "it's faster than an all-Flash array." Sure, cause there's no SAM, but it's all VM-based and brings all the simplicity you would expect from a hyper-converged platform only at scale and so what we're doing is taking that model to Linux and containers. Now, one relatively new thing we did just recently in addition to taking on VM consolidation and acceleration, we built right in all the data management capabilities you would need for back up and instant recovery, disaster recovery, archive, compliance, search, analytics, copy data management, right into the platform. So, really the virtualization guy, the DevOps guy or gal, whoever is running the applications can not only run them but protect them, share them, et cetera from one cockpit, one UI. So, we're really taking a whole load of stuff that folks have had to deal with, and tossing that for one very simple platform that scales as you grow. >> So are you bringing new services to the basic management console of Datrium and expanding that set of services across platforms. >> Exactly, that's correct. >> So talk to us about how you see this evolving as the whole world of containers comes out. Containers means, more of them, new security models. Today, most communication takes place through the VM. When you start talking about adding storage flexibility, data flexibility you guys are providing, it suggests that you've got some new ways of looking at containers. You've cooked up some new stuff. >> Craig: Yeah, absolutely, yeah. >> Talk to us a little bit about that. >> Here is where a modern platform really is important. Again, not to knock hyper-converged, but five or six years ago when that was born, it was pretty cool to manage things at a VM level, error virtualization was hot and heavy. As we move into containers, VM's are just not granular enough. In fact, folks want to be able to manage at this per container level. Arrays, we're talking about lens there. Hyper-converged is going to stop short at VM's. What we're bringing folks is a way to manage, in the VM side, VM's, V-discs, files that make up VM's, individual container persistent volumes so you can protect and share the way you need to. What we do, cause it's kind of a double-edged sword, you can manage everything at that level but now you've got thousands and thousands of them. We actually give you an opportunity to group those, what we call protection groups. Think of it as a policy group and you set it up around your applications. You set your policies per group. Through naming conventions, if you spin up a new VM or container, it's going to get included as a part of that group without you having to manually go in and assign it. So, we're effectively putting the capabilities in so you can manage tens of thousands of objects very simply. That is the world of containers, right? If you thought there were a lot of VM's, there's a whole lot more in the way of containers that will be there. >> One of the things that Datrium has done, correct me if I have this wrong but I believe I got it right, is one of the things Datrium has done is facilitated the kind of ANI addressability between storage or compute resources and data resources. >> Craig: Right. >> You know, the various of types of nodes that are in there. It used to have all the data inside of your server and that created some segmentation along those lines. In many respects, you created networks of resources that Datrium would manage in that way. Are you doing something similar now as we think about containers where you're literally describing a network of containers as part of that resource mix and being able to add things to that? Is that effectively what the group becomes? >> Yeah, the group of containers is completely independent of the servers that are hosting them so you can literally group a collection of containers across all of your Linux servers and treat that in a special way. You've got great flexibility. It's something that's really intended to scale. We've got some very powerful search tools as a part of that so if you do need to find things quickly and get it rolling. When it comes to containers, it's all about speed, keeping up the pace. Partly what we bring to the party is great data reduction capabilities, so when you're doing development in like a, let's pick on a Jenkins development environment, and you've got master/slave and you are collecting data as part of every object, all of that stuff has to move through the master. The better you are at handling data efficiency, the faster your runtime is going to be. We're observing about a 30% faster runtime for developers in that Jenkins environment, and capacity-wise, we're probably consuming 95% less capacity than you otherwise would have to do in your more traditional storage environments, so-- >> 95% reduce? >> It is a 20 to one reduction cause there's so many copies in development and we can dedupe all of that away. It's fundamentally a break-through for guys thinking about development tests, DevOps, et cetera. >> So you talked about the capacity improvements that you get in the (mumbles) improvements, but as you said, when we start going to containers, we increasingly start thinking about how fast we can add new function, how fast we can bring new capabilities together. One of the things we're fascinated about in this world, you tell me if this is a benefit that you see, is that it dramatically accelerates the entire process of doing development. Four, five, seven, 12 times speed in the development process. You not only get better runtime and do you get dramatically better utilization of resources but you are also accelerating the productivity of people that are actually doing the work. Are you seeing that as well? >> Yeah, absolutely. In fact, there are two things going on here. One is, as part of the platform, when you clone a container, you do that on your dev-server or wherever, that clone is immediately available to all other servers in the cluster. There is no copying and moving around. It is immediately available for the developers who just can go. The other interesting thing is there are, in development environments, depending on the number of developers and executors involved in development, you can have problems maintaining the state that you desire. Part of what we are doing here with these very efficient cloning capabilities, we can spin up a new environment for folks that is got pristine state which means down the line, quality is better and you're not going to thrash on those iterations on your QA cycle. From end to end, it's all faster, runtime, QA, the whole nine yards. >> Datrium's a relatively new company? >> We began shipping in February '16. We've had a great 2017, in fact, well, of course it was great. We had a wonderful fundraising in December '16, one of the largest of last year so that's really propelled us in the market. We had a wonderful set of announcements just about a quarter ago with the data management capabilities, and we added these Datrium compute nodes, and just last quarter alone, our install base which had been already showing record adoption, grew a whopping 50% in a single quarter. One of the most interesting statistics that-- >> Peter: Sequentially or year-to-year? >> Sequentially. >> Sequentially, that is whopping. >> Sequentially. The end of Q one to the end of Q two, boom. Not only that, one out of every three of our customers already has multiple DVX's deployed. That's a huge testimony to they're liking what they've got. Yeah, so it's been a sprint and like I say, we've been very vSphere-focused. Our founders are a couple of Diane Greene's. They're early principal engineers at VMware. But, customer demand, customer is king, and they're looking for the same kind of capability in their Linux and container environments so here we are. >> Hey, speed is important to infrastructure people too. >> Craig: Right on, yeah. >> So, Craig, thanks very much for joining us here on The Cube. >> My pleasure. >> Once again, great to have Datrium talk a little bit about an announcement that they did today, about adding the Red Hat environment to the great work you've been doing in VMware and vSphere, and the future of how containers, the way technology will start getting folded into that whole thing. >> Yep. >> Great results, good early start, keep it up. >> Thank you, alright, see you, Peter. >> I'm Peter Burris, good to have you once again with The Cube. We've been talking to Datrium about their new announcement. Craig Munes, er, Craig Nunes (laughs). Craig Nunes of Datrium, Vice President of Marketing, thanks for being here, Craig. >> Craig: My pleasure. (techno music)