(upbeat techno music) >> Live from San Francisco, celebrating 10 years of high-tech coverage, it's theCUBE. Covering VMworld 2019. Brought to you by VMware and its ecosystem partners. >> And we are back at VMworld 2019, here in San Francisco, along with Stu Miniman. I'm John Walls. Welcome to theCUBE here, continuing our coverage here at Moscone. And we're now joined by Don Murawski, who is the manager of Servers and Storage at Wendy's. Don, glad to have you on theCUBE. >> Glad to be here. >> Before we leave, you're going to have to settle this with Stu. He's very upset about the change in menu. That number eight is no longer number eight. >> I can't do anything about that-- >> Well, perhaps we're going to look into that a little bit later on. And what you can do something about is is tell us about your portfolio of services. What you do in terms of, what you are managing in terms of storage, in terms of servers at Wendy's. >> Right, right, yeah. As you know, IT has, especially in the food industry, has become huge, especially mobile app, mobile ordering. You know, DoorDash. Order your food and have it delivered to you. You know, massive business. Massive financial for the company too. How that plays for me is, managing the infrastructure that it runs on. Whether it's an AWS or Azure. A lot of that stuff is on-prem still. So we have to manage a huge amount of volume with a very dense environment. For a hyper-converged shop, Nutanix is part of that. Cohesity is a huge part of our system now. Two data centers. One in Atlanta, one in Dublin, Ohio. So, it's quite quite a big effort. >> You mentioned on-prem, off-prem. About what's your split right now, and are you-- >> I'd say 30:70. >> Okay. >> Yeah, 30 off, 70 on. >> And how is that going to change, you think, over the next three, four, or five years? >> Oh it's changing, drastically. Cost. You know, CapEx, OpEx. It depends where our model's going to be at. Right now, we're more CapEx. So when that goes to OpEx, you'll see a lot more cloud. So right now, 70:30. 70 on-prems, 30 off. >> All right, Don, we talked to so many companies today, and what is that digital transformation they're going through? You talk about app and mobile. It's like boy, I'm reading articles about, how do we make sure that your food delivery person, doesn't eat a lot of french fries, before it gets to you? Maybe speak a little about the ripple effect that has to, your group and IT, as to, you know, we always say fast food. What's faster than walking up to the counter and you know... You guys don't have it sitting under the warmers, of course. They put that together and make it. But it's now transforming that fast food business. >> Yeah, it was touching the back-end servers So it's important that those are properly tuned, properly functioning, on legacy, sometimes legacy hardware. So between cloud and on-prem, it's been a challenge. And we're still working through that challenge. A lot of our developers are in-house. We actually have a big presence for developing right now for our own app. We actually develop our own app and websites. So a lot of that is tied into the movement of, more into cloud technology, than on-prem technology. So right now, like I said, it's 70:30. But it's still a challenge. >> And what is it about that when you say it's a challenge, I mean. So we've drilled down on that a little bit. >> It's just dealing with, not (mumbles) With on-prem you can't scale like you can with AWS or Azure. You can scale 100 times down an Azure bot, auto-scaling. On-prem we can't do that quite yet. We're getting there. So that's still a challenge, because a lot of the information still hasn't touched, on-prem. On-prem databases, which are getting older too, so to speak. So it's still a challenge. >> Don, when you talk to companies, you talk about that whole modernization. And the keynote this morning. We're talking about hybrid-cloud. We talked about multi-cloud. HCI is often a piece of that modernization, but how do you look at how you scale and change things in your data center, versus the public cloud. Is it making progress? Is it limiting at all? >> It's slow progress, slower than we want. More like into, getting rid of the VMs, go containerization. That's a lot of containerization that's happening now with Kubernetes. We have a DevOps apartment we actually just created internally to do that type of work. It's just taking a little bit longer than we anticipated. >> Yeah, and (mumbles) obviously Kubernetes is big discussion here. >> Right. >> How long has your group been using it? >> Not a year. >> Why do you use it? What is it? What's the value to your organization? >> Click a button and you've got a server. It's auto-scaling. So instead of taking two hours to build a server, or three, it's taking two minutes. I think we actually timed a Linux server build in two and a half minutes. The fact that you've got a small workforce too. I mean, we're advertising jobs. Things are what they are. They're pretty stagnant. So we have to make do with the technology that's out there. And Kubernetes is a big part of our future, infrastructure. >> But oftentimes Kubernetes is something that will help me if I want to move something from my data center to the cloud or between clouds or like, do you use that use case yet? Or -- >> Not yet, not yet, we're getting there. >> Okay. >> Yes. Slower progress than we'd want, but yeah, we're getting there. >> All right, when you're living in this multifaceted environment, bring us inside your data management. What's that like today, what's working, what challenges do you have? >> I'll tell you what it was like. It was a nightmare.(laughs) >> Yeah. That'd be awful. (laughing) >> It was a complete nightmare. Multiple vendors. Very complex. Now we're trying to simplify things, make it more dense with (mumbles) Cohesity. It's been a big part for the past year. We moved all our backups at Cohesity. So Cohesity is basically backup and DR now. I don't use it like secondary storage. I have other storage for that, smaller storage units. So it's... Two years ago, we had lost our primary storage and basically took down the company. And living through trying to get your data back for hours and hours and hours, and working. I had guys working 100 hours a week for two, three weeks. And (mumbles) didn't see their families. So making something that is easy to use, manageable, and recoverable, was huge. So take the complexity out and add the ease administration. And that's what we did with Cohesity. >> Yeah, and you're painting really maybe not a worst-case scenario but an awful-case scenario. >> It was an awful case scenario. >> Yeah. So I mean, disaster recovery was a disaster for you. It sounds like that. >> It was. >> So is that what drove you to the Cohesity decision? >> It was, that was a big factor. The fact that I need to be able replicate this stuff to another location, that's one thing. That's what everybody says. But can you actually recover it if it's in the other location. No, we couldn't. Now we can, and I actually prove that through a POC. So yeah, it was a big factor. The fact that people had to sacrifice weekends and I mean literally, work all night, multiple nights, to get things back up, to get the business back up. >> So what do you say to your colleagues or counterparts out there, maybe who haven't, maybe done this kind of spadework that you guys from -- >> I'd try to turn down your critical servers and see if you can recover 'em. You know, take them down and see if you can get 'em back up. Test your DR, because if you don't, it's going to come back to bite ya, and it did. We got most of our data back, but there's some things we didn't get back. We had to recover, I think we had to retire a couple systems that were homegrown systems that were written by a developer back in the day, that's no longer there, we couldn't get it back. We had to send whole departments home, because of this. So I would say, test it. Make sure it works. And make sure your vendor, whoever you pick, is standing by you too. That's a big thing, it's that relationship with the vendor. We don't pick it because it works, we do. But we also pick it because of the relationship with the vendor. Are they going to be there when all, you know what, breaks loose. >> Right. >> Some do, some don't. >> Who's your friend right? >> Who's your friend. >> So Don, you've gotten your title. It's Server and Storage. But you're talking about the Kubernetes, and modern multi-cloud environment-- >> Don: We're small shop. We're small IT shop. So out of my group, DevOps actually spun out. So now we're kind of a infrastructure DevOps team That DevOps is a whole separate thing now because of my team. >> Yeah but (mumbles) what I guess I wanted to get it right is that, was that mostly internally training and going through the model. >> Yeah, it was. >> Bring us through some of those, what worked well, what was a little bit of a pain point. >> Pain points, It took a year to get one application working. But now it's working, you see the value in it. Because I was like, this is a waste of time. We don't scale that much. But however when you do, it sure is nice to build a server like I said, two minutes, that's a huge factor. You know, it was coming, that seemed to be the trend. DevOps, I mean, we wouldn't have DevOps jobs three, four years ago. Well now there's DevOps admin jobs. So it was coming, it was just a matter of time. >> You've been using Cohesity for about a year now, you said. >> A year. >> You talked about where you're using it. Give us a little bit looking forward. Where do you go with Cohesity. What would you like to see them do. >> Yeah, I think a big point is going to be, especially from a (mumble) infrastructure platform, will be more of an Azure footprint. Shrinking the on-prem data center. So Cohesity is going to play a huge role. We still have a lot of 2008 servers. And 2008 goes out of, end of life, in a few months. There's no way I'm going to retire 200+ servers by January. It's not going to be humanly possible. So a lot of that stuff I had to get moved to Azure for support, and Cohesity's going to play a big role in moving that and protecting it. So yeah, I'd say a good path for Cohesity in the future for us. >> So when I brought you on and we talked about the menu, item number eight, that you said you can't help Stu with, is that right? What is menu item number eight? >> It was one of the chicken specialties (mumbles) they have. >> So however, if that, we are very supportive. >> Don: I like that. >> We have our $2 Frosty donation for the year. >> Don: It's good. >> So a Frosty a day right? A free Frosty a day. >> That's right. >> So, we are supportive. >> That's good. >> If you can work on that number eight, maybe-- >> Don: I'll work on it. >> Maybe we can be even more supportive. >> All right John. >> Thanks Don. >> I appreciate it. >> Absolutely, this belongs to Gabe Leon, by the way, on our crew. Just got to give Gabe a shout-out there, for helping us out. Back with more here on theCUBE. You're watching this live, VMworld San Francisco here, 2019.