 Okay, we're back. This is Dave Vellante of Wikibon.org. I'm here with David Floyer also of Wikibon.org. This is SiliconAngle.tv's continuous coverage of EMCworld. We're live in Las Vegas. This is theCUBE where we bring you the perspectives of practitioners, CEOs, CIOs, executives, bloggers, analysts, the smartest people in the industry. We extract knowledge, we share it with you, our community. Thanks for watching, everybody. This is the future of data protection spotlight that we're running here at EMCworld. It's a drill down into data protection. We've been looking at a number of issues. How virtualization is affecting backup. We're looking at the future of backup. Where CDP continuous data protection fits in. And we're here with Bill Holmes, who's the VP of IT Infrastructure at Loomis to bring the practitioner's perspective. Bill, welcome to theCUBE. Glad to be here. Thanks very much for coming on. So why don't we start? Tell us a little bit about Loomis and what your role is there. Loomis, we're an armored car company. We pretty much handle cash in transit. We handle cash management. A lot of the banks now are outsourcing a lot of their cash management services. So they're sourcing that to us. We have a lot of transmissions that go in and out of our company. We also have a safe product, electronic safe product, which we are in a lot of the retailers and we actually pull throughout the day. So we create a lot of transmissions of the banks. We provide provisional credit. And because of that, we have a pretty good IT department and it's kind of grown over the past several years. So that's kind of where we are. My role there is really just over the infrastructure and making sure everything's running and we're meeting our SLAs. So you're all about protecting your business. It's all about protecting people's cash. Yes, exactly. Exactly. So you're probably pretty stringent with how you protect your data. Yes. What's the mindset? Well, we don't want to lose anything. Like I said, we get audited by the banks on a regular basis and we have to make sure that their data is being protected and we have to show that it's being protected and we have to prove it. So we go through audits every few weeks from all the various banks. There's over like 280 something banks that we deal with. So a lot of process around that. Yes, exactly. You sort of prove that you've got your act together. Right. Essentially. And that's, I would imagine has evolved as that industry has changed. Yeah, and it has. And I mean with the cash management side of it, we have 88 branches that are out that do nothing but handle a lot of the cash management, which each one of them has their own server in it. And that's kind of where we came in to play with the EMC and the BRS backup side of the house with Avomar. So before we get into the whole Avomar piece, just set up your shop for us. What's the infrastructure look like? Servers, virtualization apps you're running? Give us the rundown. What about? We add the story as to why you had a problem in the first place. Sure. What made you look at some things in a different way. Okay. When I first came on board about five years ago, it was a very physical environment. We are now at 75% virtualized using VMware. We've got a majority of our servers completely virtualized. We run SQL, Oracle, we have a Cisco network. So we've got about 500 servers that I manage on the production side of the house and probably about 300 servers on the Dev and Test side. So we've got a pretty good size network. Like I say, I have about 80 servers out in the field which handle nothing but our cash management piece of it. So every night they send transmissions back to us and we turn around, rebundle and package them and send them out to the banks every night. So it's pretty intensive. So at what point did you decide to go with that? I presume like many customers you started to virtualize. Right. And now your backup windows are getting squeezed. Yeah, we were actually hitting a 24 hour backup window. I mean, it was getting ugly. We were having to, it was killing our production, yes. Exactly. It was killing our production. It was taking our network down. It was pretty rough. And the bad thing is what really started it was with the 80 servers that we have out in the field, we were relying on the branches to do the backups. Local tape. And we actually had a server fail and we tried to recover it and we had a major problem trying to get that thing recovered. At that point I said, we got to come up with a better solution and that's when we started engaging our partners. I heard on the keynote that somebody forgot to change a tape for seven months. Seven months, yes. That tape was stuck in there and having a tape in the tape drive being written every night just kind of wears it out. So it was done. Yeah, so you made the call to basically essentially change your entire data protection philosophy processes. Yes. Everything around it. Yep, we got the Avamar piece in but at the same time, I mean, just to, backups were just now beginning to emerge and that was kind of the direction I figured we had to go. I had, throughout my entire career, we've had to deal with tapes. Tapes don't work all the time. You try to, I tell anybody, you try to restore a server from tape. You got a 50-50 chance that it's going to work. So yeah, it's been a nightmare. Yeah, so, okay, so I guess in one respect, the decision wasn't that hard, but at the same time, you had to endure some pain when you migrated from whatever environment you were before or whatever backup application you had, you had to essentially rip and replace that, right? Yes, we did. And honestly, it really wasn't that bad of a problem. I mean, everything went in fairly smooth. We started, like I said, we started with the Avamar piece just to get everything from the field in. After that, we started with the data domain, the networker, we started bringing all the pieces together. And each time we kind of kept lowering our backup windows and lowering our backup till we got rid of all of our legacy-type backups. Did you get the support from management to invest? Were you able to, because of the accident, were you able to then say, well, because of the accident? Yeah, the first accident. This is kind of important. Yeah, yeah, it was really easy to get management by. And I basically, I painted a picture of Total Disaster. And the banks, like I say, were coming down and they were doing audits. And we needed to have an off-site storage and everything else. So we were able to use the replication of the Avamar and the data domain to replicate off-site. And that satisfied a lot of the banking requirements of having an off-site storage. So it was really a customer service driver for you, as opposed to any kind of internal issue. There was some internal stuff, obviously, but it was really triggered an external, whoa, V8 moment. Yeah, with the customers and the banks, that was the biggest thing. So now, I think I heard you say Avamar for the source side duplication, presumably. And then you've got data domain as well. Yeah, we've got data domain using as our VTL. And we're using NetWorker as the backup software. It's okay. So you've got Avamar for your remote offices and the NetWorker. And we're also using Avamar to back up our VMware. So your VMware and then your big database environments and other things you're using, the more traditional. Yeah, the data means acting as a VTL. So we're backing up to that right now. So what do you see is the future? We had Stephen Manley on here and he, I don't know if you heard his keynote. Yeah. And he put forth this vision. We've been talking to him in previous events about we've used this analogy of the time machine for the enterprise, you know, the Apple time machine for the enterprise using snapshots and CDP. Is that, in your mind, a viable vision? Is that the direction that you're headed? You know, data protection as a service? Is it too early for that? What's your perspective on that, Bill? No, I mean, that is where we're headed. We've, in fact, we were looking at doing a refresh probably around the first of the year with our entire backup system again, taking it to looking at some of the faster processing that they now provide on the backups. We went, because we went from a 24 hour backup window to the new solution, which took us down to about an eight to 12 hour backup solution. And I think with some more efficiencies, I can get down four or five hour backup. So at that point, we've got some really strong efficiencies. And it gets to the point where we get to refresh our entire backup system, especially with some of the new technologies they're offering. Why did you initially choose to go with EMC? Are you a predominantly EMC shop? We were on the storage side. And we had a pretty good relationship with our sales team. My philosophy is I like to partner with the sales teams as much as I can. And that was one of the things we brought them in with that first problem. And they basically came in, dropped in a PO proof of concept system. We ran it on a couple of our branches. The deduplication was fantastic. And I said, I can't go wrong with this. So that's when I, we took it to the next level. And eventually we just continued to keep growing with it. Everything we put in pretty much worked. When you say the deduplication was fantastic, you mean the rates of deduplication? The percentage, we started getting up around 97, 98%. So the amount you had to transfer over the network was pretty small? Very small, yeah. We were getting down only two to 3% transfer of data. So yeah, it made it a lot more efficient. And with the VMware, everything else. I mean, the stuff doesn't change that much. So you're backing up a relatively small piece of it. So. And are you replicating your target data offsite? Yeah, we have a Avamar and Data Domain secondary node. Everything replicates offsite every night. So we keep offsite copy. Excellent. What about your long term? My goal event, well, right now we're having to tape out about every three to four months. So we're still sending a little bit off to tape. My goal is to give it a tape completely. By this next year. And that's why we're looking at refreshing our Avamar and Data Domain PSTS drive in it as well. So to clarify, getting rid of tape for backup or getting rid of tape completely? Completely. No tape initially. No tape, yeah. And your clients are cool with that? The banks are okay with that? Yeah, in fact, I think a lot of the banks are doing some of this stuff already. But yeah, they're fine with that. As long as we're replicating it offsite and we're able to retrieve it, we're there. So talk about that retrieval process. So if you're in a disaster recovery situation, how do you get it back in a timely manner? Do you fail over to that other site? Yes, and we also have servers there that we can rehydrate the data back and bring the servers up fairly quick. Right, okay, so you don't have to move the data? So when you rehydrate the data, you do that remotely and then transfer it? Yeah, well, we can do it from either node, either the primary node or the backup node. We can rehydrate that to a virtual box and restore the servers. One of the things we do every couple of months is we actually test out our backups. We restore several of our servers, we keep logs, making sure that they're tested. And usually we just restore it to virtualized servers. And so far, it's been working pretty good, no problem. Testing things, all the concepts, isn't it? Yeah, and the banks love that. Okay, so we're here at EMC, well, should we make this whole IT transformation in cloud, big data, transform IT business in personal? Is that transformation something that Loomis is going through? Yeah, we're definitely transforming through that now. We're building out our private cloud and looking at our DR sites, replicating, you know, being able to move our cloud to any site that we need to go to. So VM, yeah, the virtualized side of it, the data backup, everything. We also use another EMC product, MoziPro, to back up all of our laptops. Any of the laptops that are traveling, they get backed up every night into the cloud. So in regards to that transformation, is it predominantly focused on the infrastructure or would you say it's really more toward service catalog, IT as a service? From my standpoint, it's probably more from the infrastructure side. I really don't see it so much as a service so much to the business that they don't see it, they don't really know it. If anything, it's more of an IT infrastructure piece of it just from a management standpoint, from a control standpoint, and being able to do our due diligence. You guys don't do charge bags, do you? No. No, we've looked at it, but no. It's a size of company where you can still manage IT that way. It is, we're a global company, but we only manage the United States, that piece of it, North America, so. Have a local. Yeah, so, and each of the other countries manages themselves in the same manner, so. Good. All right, well, we're here with Bill Holmes, who's the VP of IT infrastructure at Loomis, Loomis US, Headquarters based in Houston. Yes. All right, European based company. Bill, thanks very much for coming on theCUBE. Thank you for having me. And sharing the practitioner perspective, I always like to hear your guys' angles and keeps us honest, so good to meet you. Nice meeting you. Thanks very much, appreciate your help, David, and thanks for watching, everybody. Keep it right there, we'll be back with Jason Buffington and David Flora, breaking down the future of data protection from an analyst perspective. This is theCUBE, SiliconANGLE.TV's continuous coverage, live from EMC World in Las Vegas. We'll be right back.