 Live from Las Vegas, it's theCUBE, covering Dell Technologies World 2018. Brought to you by Dell EMC and its ecosystem partners. Welcome back to Las Vegas, everybody. You're watching theCUBE, the leader in live tech coverage. My name is Dave Vellante. I'm here with my co-host, Keith Townsend. Craig Nunes is here. He's the CMO of Datrium. Yeah. Long time Cube alum. It's great to see you again. Great to be back. Awesome. James Stock is a Datrium customer. He's the vice president of IT at Grow Financial. James, welcome. First time on theCUBE. Looking good, man. Yes, thank you very much. All right, Craig, Datrium. You guys smoking hot, changing the storage world. Give us a quick update and we'll get into it. Look, we are, we're filling a huge gap. Bigger, I think, than we had imagined because look, a lot of, it's no secret, the array market is in decline and hyper-converged has tried to kind of reinvent that market and it has to a degree on the low end, VDI, that kind of stuff. But data centers need an answer that scales. They need an answer that's got resilience and it turns out, after all these years, backup is still a problem. Figuring out the cloud is still a problem and so we put together a system that really takes a tier one approach to HCI, a full-on scale-out backup system and a cloud DR approach built into one converged system and customers love it, from cloud to backup to performance in primaries, it's been awesome reception. Well, let's see if they really love it. Yes, yes. So, James, first of all, so let's start with Grow Financial, your role. You heard the pitch and then we'll get into sort of how you're applying it to your business but tell us about your company. So, we started in 1955 in a broom closet in a MacDill Air Force Base headquarters there in Tampa and over the years we've grown. We're now $2.4 billion in assets. We have over 200,000 members and we do lending throughout the Southeastern United States, offices in Tampa and in South Carolina. So, and your role, head of IT? Basically, what I tell people is that if it plugs in, I'm responsible for it. Okay. All right, so take us through the sort of day-to-day project of before and after. What was the motivation? So, really, I mean, the issue that we were running into is that our existing storage solution, which was the Dell SC, was, our trays were running into life and if we only had a couple of them, probably wouldn't have been a problem. We might not have even entertained it but we had probably two dozen. So, we started looking around and I said, all right, well, what does it cost to replace what we got and what else is on the market? And we started to find out that just replacing what we had with like was gonna cost almost 200 grand more than what our full day-to-day replacement cost. So, it started making financial sense right away. But I think what I, we met up with day-to-day probably kind of been summer of 2016 when they were on version one. And it looked good. You can see the promise, the whole idea of having that back-end storage that was really intriguing because none of the other players had anything like that at the time. And we said, all right, we're not ready. And then when they came back out in May of last year, whoa, the difference in what they'd done in such a short period of time is what really kind of blew us away. Okay, but we're here at Dell Technologies World but you guys are a partner of Dell's, right? So, you're using Dell servers and, right, that's part of the deal here. So, they let you in. Yeah, Jason, they let us in. In fact, our compute nodes, it's no secret, are Dell branded compute nodes. And in fact, we have partnered with Dell in one of their data centers to set a world record IOMark on Dell Gear just to prove a lot of the performance specs that we've shared in the market, prove it out and we proved it out on Dell Gear. So, James, talk to me a little bit about your perception of open converges. I've talked to Craig about open converges versus hyper converges versus converged infrastructure. At the end of the day, you just want a reliable, fast system. However, what about the open converges story drew you to data? So, I didn't have to replace any of the nodes I had if I really didn't have, if I wanted to. So, I've got Cisco nodes around my call center. I've got Dell nodes, got Datrium nodes now but at the time it wouldn't have mattered. I could have just, like in my Cisco environment, I actually had to add a RAID controller to the UCS box and then I could throw any solid state drives that I wanted into the device. So, that was where it really got compelling. I'm like, wait a minute, so you're telling me I don't have to buy enterprise-class drives and stick these in each of my servers. I could just go down to Best Buy or wherever, local, grab some off the shelf and throw it in there as long as the server supported it and, okay. Where do I sign up? So, we've heard that story and one of the things that some of the other or some of the hyper-converged infrastructure players say, you know what, we could do that but it's almost impossible to support because of firmware issues, et cetera, et cetera. Did you guys run into any of those issues? Nope. That's been the greatest thing. When we first started to do our reference calls, I was like, everybody I talked to, I said, well, where's the catch? Right. It seemed too good to be true and customer after customer that I called, they said, you know, like we ran into it with our backups but they finished like a third of the time faster. I said, how is that even possible? They said, we didn't believe it either. We actually had to go back and check because some of our backup jobs finished so fast we thought it was an error or something like that. You thought they didn't complete. They were fine. It's just you're backing up from Flash now instead of backing up from old spinning disk. Okay, so you put the system in, talk about the business impact. It sounds like there were some residual impacts from the initial motivation. Right, right. So from a business impact, that's a tough story to sell because really where we saw it was on the back end and that was the way our systems were before. There really wasn't a huge deal of impact to the business with our old system until it came to backup times. Now where I will say that we saw reductions is if I have to reboot a server today, so our call center application, prior to putting it on Datrium, it took anywhere from 15 to 20 minutes for that to boot up. Well, 15 to 20 minutes while our call center's down is like an eternity. Now that time's down to about five to seven minutes. So like overnight, you've more than half that time. It's the same thing with web servers. Or anything else that would be member facing, those times have been greatly reduced. So if I do have to reboot something because everybody knows it happens, it's sped up the process tremendously for us. And what's the secret sauce here? We're talking architecture, just sort of modern approach, software design. So the secret sauce, if you will, is this split design that runs your workloads, especially your read intensive workloads on Flash, on the host with powerful software, the Datrium software. All of your durable data does not live on those hosts. Those hosts are not stateful. They can fail at any time and you still have data availability. So you've got that bullet proof availability and on the back end, your data's kept secure. It is shared so we don't have any network traffic between hosts, your network doesn't blow up when you install like it does with a hyper-converged approach. And that split provisioning, that split architecture, is kind of the breakthrough. And that's why we talk about beyond HCI. We took kind of the good stuff there, the scale ad attributes, VM, centric admin, but then we really built in tier one capabilities full on backup. And of course we haven't talked about it, but access to AWS for your off-site backups. So Jay, let's talk about day two operations. One of the advantages of hyper-converged is this idea, I have one pane of glass, I can manage firmware updates, I can streamline my operations. Do you guys see similar advantages, day two versus your previous infrastructure? Yeah, I mean, one of the things that saves us a lot of times now is the fact that there's just one big pool of data out there. Instead of having to provision LUNs, like we were setting up our exchange conversion. So we're building out four or five servers for that. Well, normally that'd be about a two-hour process, not that we were sitting there waiting the whole time, but all right, we'll carve out some space in this LUN. So what are your thumbs, go do something else, come back, and maybe that would be done. Well, now that's like an instant process. So those sort of things are like, wow, I'm saving tons of time just in admin experiences. In terms of pane of glass, it is a single pane of glass. One of the cool things that we run into is, every now and then, of course, we've got to do our disaster recovery testing. We're a financial institution. Well, Datrium's approach is really unique and a problem that we used to have is if I failed over to our DR facility, well, now I got to bring that data back. Because when you're failing over, it's not a problem. You've already seeded that data. Well, it doesn't work the other way around. It does with Datrium. So with Datrium, if I go to bring that data back, it's not doing a differential copy back. So I'm not sitting there for days and days and days waiting to finish my DR testing anymore. So there's just so many different benefits that have just, have been great for us. I mean, that's huge, because a lot of times organizations, they can't test DRs too risky or they just don't have time in under the resources. Right. Did you have that problem beforehand? Or did you just? Well, yeah, because what you would run into is that it took so much time to do it before that I'd had to run my guys ragged for two or three weeks. Like, all right, stay up overnight. Make sure it all copies. And then once it's copied, okay, bring it back up. So those, yeah, I mean, that was a challenge before that it's not a problem anymore. You're burning the team out, right? And or missing your windows. Well, and because of the way that it's architected with the protection groups, I no longer need to use a third party recovery tools to do the transition back and forth. I can do that natively inside of their application. I always like to ask practitioners, you know, if you had a mulligan, what would you do over? And often it sounds like nothing, but or what kind of advice would you give to your peers embarking on a similar journey? Do all of your reference calls, see it for yourself. I mean, I take quite a number of reference calls because people are in the same boat I was. Is it true? Does it really work the way that you say it does? Yeah, it does. You know, I'll screen share with them. If they want to see our numbers, I'll show them. All right, last word. What are we looking for? What are we looking for? Looking forward. So you're going to see us double down on the work we just put into market, our DVX4.0 software, which comes with that cloud DVX, cloud based capability, and take that into full on disaster recovery, orchestration, and not in the two distant future. You'll get the whole rundown, so speak to. Awesome. Craig, thanks for coming on. James, pleasure meeting you. Likewise. Thanks for everything. Thanks for hanging out with me. Always great. All right, Keith, good job. Good questions. All right, keep it right there, buddy. We'll be back with our next guest right after this short break you watching theCUBE live from Dell Technologies World 2018. We'll be right back.