 Okay, we're back, this is Dave Vellante, and we're here live at EMC World 2013 in Las Vegas, amazing show, 15,000 plus people, we've had all the executives, this is the Cube, Silicon Angles live production of EMC World where we extract the signal from the noise, we bring you the best guests that we can find, we like to talk about tech athletes, and we've got a tech athlete with us, Mike Summerville, who's with the University of San Diego, and we're going to unpack some of the discussion that we've been having around converged infrastructure. Mike, welcome. Thank you so much. Pleasure. Great to meet you. So, tell us a little bit about what you do at USD. So, my title is Manager of Systems Support, and I'm the Chief Cloud Evangelist as well, and my team supports four data centers on campus. We have a bunch of systems architects and network systems architects, and so we just basically support the whole infrastructure. We have a converged IT department as of about 2005, and so we provide all the services for the academics, the administrative staff, and everything that's needed to run an enterprise of about 8,000 undergraduates and about 2,500 employees. So, what's the IT organization and the infrastructure look like? Maybe you talk about the infrastructure, the apps, you know, how you organize. For sure. So, the infrastructure, our network is a Cisco-based network. We went with the Nexus 7000 series, and so we have a 10-gigabit network, fully on campus. We have a wireless cloud over the whole campus, provided by Aruba, which is made up of about 1,850 access points. We have 802.11 and ABN, and so at any given time, we have about 1,750 concurrent students on our wireless. We have gigabit pipes to the internet, like most universities. The real success we've had, as you brought up, converged infrastructure, has been with, we started out with UCS, but then recently with the VCE, the VBlock solution. And I apologize for being late, I was actually over at the VCE, but I was talking to those guys, trying to get the word out there about VBlock and how much of a great solution that's been for us. It's one of those things that we can just order, roll into the data center, literally within four to five days, turn it on with some IT rock stars, like I mentioned, and be up and providing exchange and SharePoint and all the other services. So before we get deep into the VCE and VBlock piece of it, I just want to understand your environment a little bit more. Talk about the applications that you're running. Okay, so for email, actually, we have all of our students out in Google, but for on-premise, we have Microsoft Exchange, we have about 3,500 accounts in there. We run SharePoint, it's one of our portals. Our student system is Sunguards Banner, and then we have their Luminous Portal to interface with Banner and the student system and the grades and all that. We have other enterprise applications, as you might imagine, so Hyperion and Cognos and the like to do. We have Oracle e-business suite for HR and financials, Chronos payments, or Chronos time card system, and a Microsoft point of sale system, and a lot of that's virtualized. Okay, I was going to ask you. So what is not virtualized? So great question, actually, and it's funny because I talked to a lot of people and they're like, well, aren't you 100% virtualized yet? And I don't ever see us getting there. I actually think that being 95% virtualized, 96% virtualized where we're at right now is a great spot for the University of San Diego, mainly because the stuff that's not virtualized are appliances. Like, Google doesn't give me a virtual appliance. They give me a physical cool yellow box. And so there are other things that we do network monitoring with and those come as appliances, physical appliances that I install in the data center. I've got a lot of holes now in my racks because I've got a V block and I've got virtualization and I've got UCS, but for the most part as far as I'm concerned, we're 100% virtualized as far as our goals. But your Oracle is virtualized? Our Oracle is virtualized. We use the IBM system for that. So we have AIX, so they're LPAR technology. And we adopted that several years ago when version five just came out. So now we have IBM 770s and that supports Oracle e-Business Suite, all of our Oracle databases. So prior to rolling in the V block, you were quite a ways down the path of virtualization. We were actually, we hopped on the bandwagon around 2005, 2006 with the proof of concept. And then in 2007, we really started rolling it out when 3.5 and 4.0 started coming out for VMware and said, okay, this is going to be, and it was a classic discussion. It'll be great for Dev and test for write, you know, but it'll never do production workloads. So I went back and forth to even two years ago and it's like, if I don't run Exchange in a virtual environment, A, I'm a moron, secondly, it's not going to perform as well. You know, I remember when I first heard, yeah, Exchange runs better in a virtual environment. I said, wait a minute, does this say VMware on your T-shirt? And as it does, it runs better in a virtual environment. It's fantastic. So I want to unpack that a little bit because the application guys are always saying, well, you're putting in some kind of layer and there's a tax. Well, their application is a layer, let's be clear. So if they want less layers then. So that's a myth, right? This day and age is a myth that that layer is going to cause performance problems for your critical application. I absolutely agree with that statement. So why the perception and how is it actually better performance? So I think as far as the perception goes, I just think it's lack of education and potentially lack of experience. I know that if I were to roll out virtualization and I wouldn't put it on decent enough hardware, it would be a performance problem. But if I did a bare metal install on hardware that wasn't performing well, there would also be an application problem. And so I think that's the first thing. Secondly, I think all we can do is just roll out more and more applications the same way and people just start getting more comfortable with it. Of course it's virtualized, of course it's virtualized. And it's only now that we're looking at it saying, well, instead of me doing all of that myself, maybe I should buy the converged infrastructure, maybe I should buy it in a box, okay, so you decided to bring in a V block. What was the motivation behind that decision? I mean, you were running pretty well. You had a nice virtualized environment. You know, why the V block? For sure, so actually we had kind of done our own V block that I hear a lot of people doing. So we had a UCS chassis, we had EMC storage, we had VMware and we had a Cisco network, 10 gig all over campus. And so we were pretty proud of ourselves. We said, hey, we did a really good job with this. But is that the business we're in? The structure that supports that experience to allow our students to be able to just pretty much do everything in the sense that the technology is transparent to the user it serves, you know? And so we've been able to do that very successfully and we said, all right, well so who's making these things? Who's making this stuff that we have in our data center now? And there were options, you know, FlexPod and that sort of thing, but VCE came in and they said, well we've got all the stuff you've got, Mike, and in fact we wrap it around this wonderful black box and we roll it in and we do all these things. And I said, well that's great, now we can actually be creative and we can make our applications maybe run a little faster with some tuning. We can do some Java tuning or whatever. Okay, so was the timing just good? Were the assets depreciated on the books or did you have a new application that you were rolling out? Again, what was the catalyst to bring that in? The short answer is we needed more capacity. So I could either, we had not depreciated the assets, we were only about a year and a half into our UCS sort of marriage that we began. Our NS40 that we had was only about 18 months to two years old. And we just needed more. And I said, okay, so do we buy more UCS, do we buy more storage and just keep throwing it in or do we make a paradigm shift and adopt what Cisco had been talking about at the time with just the converged infrastructure and bringing the network outside of the UCS chassis, right? And so we went with that route. So you saw it, made sense, you bought into it. What were your concerns before you rolled it in? Well, my concerns were that it was going to be too expensive, right? I thought, well maybe there's going to be a soft sell afterwards. Like, you know, Mike, now that you've got this thing. Oh, you want to turn it on. Yeah, exactly. Oh, you need someone that knows how to use it, you know? And so, but it was great. Included in the price, of course, was, you know, on-site deployment as well as some hand-holding, you know, and some knowledge transfer and some training so that we would be successful with it, right? And any mature company is going to be like, well, of course we're going to train you on our products so you can be successful with it. Otherwise, you know, why are you putting it in your jade center? So what apps did you put on there? The short answer is all of them. So everything that we had virtualized, we had about 170 VMs. So Exchange, SharePoint, Hyperion, Cognos. Any of the time card systems are E-Online Learning System, which is WebCT or Blackboard now. And then every file share that we had connected back end to either iSCSI, NFS, or CIF shares that we present to the faculty. All of AD, LDAP, every, you know, production system. So Mike, talk about the experience of bringing it in. How long did it take? You know, what was involved? So I know this is going to sound like I'm working for VCE, but how long it took was it showed up on the dock on a Friday. All of the guys arrived there on Sunday afternoon and then we said hello and then we went basically and we installed the thing Monday and by Thursday we were spinning up VMs. Really? And if you think about it, I mean, it's not like it's a pie in the sky idea. I mean, again, they built this thing ahead of time. I kind of expected it to work right away and they exceeded my expectations. We had a team of about six guys on site and we had a network guy and a storage guy and then we had a VCE guy who got the whole thing. We had a virtualization guy and they all worked in our data center. We had a table just like this set up and then we went back to one of the classrooms that was nearby our data center. We whiteboarded it all out and decided, okay, so what loads are we going to put in this? How are we going to carve up the storage? We're definitely going to do RAID 6. Are we going to do N plus 1? Absolutely. How are we going to do that? What are the notifications we're going to put in place? What are the monitoring systems that we're going to add in? All of the other things in your environment, you've brought this big V block in, clearly you get virtualization. So you have some VMs running somewhere, like almost 200. How are we going to get those inside there? And so then we extended the fiber channel from just inside of the V block to the whole infrastructure and then the storage V motion over about 16 hours and we were up and running. All right, so break down the metrics. What was the business impact? So the business impact was just simply that the users get their services faster. I don't have an exact number of hours that it used to take versus the new one, but I can tell you that we've almost shot ourselves in the foot because we can provide VMs so fast now that people are like, well, Mike, I'm going to need six VMs on Monday and that's fine for a while, but then I eventually run out of capacity. And unfortunately, I tell you, the only reason I'm kicking myself is the fact that I just didn't buy enough EMC or VCE stock because what I thought would last me three years, it only lasted me one year. And it was interesting at the VCE users group the other night, I had four other guys say that, the guy from Visa, the guy from this gas and electric company, they said, yeah, in one year, we used up all of our capacity because we now have the ability to turn out these VMs that are incredibly reliable in a load balance configuration very fast. And so the customers are like, great, bring it on. So, time to value was the real main driver. I would agree. Now, if I understand it right, it's not like you redeployed resources because you got, you accepted greater demand. You increased your capacity to deliver more services. Definitely. Okay, so now, how did that all work within the organization? Did the organization notice? Are they giving you more money? You now have to figure out how to pay for the future. Yes, they're giving us more money. I'm grateful to work at the University of San Diego for a number of reasons. One is that we're small enough and nimble enough where we can turn around POs and actually get the ball down the field a lot faster. Additionally, they understand IT and they understand that while IT, certain things, the speed of something maybe gets cheaper. The amount is growing. We're not at the point where we're actually talking about big data at USD, but we have a lot of data. And they understand that there's a need for flow there. I think the best way to describe it on campus is that no longer is the IT infrastructure team the bottleneck for getting a project finished. The developers are able to do what they want to do and we're not slowing them down. Great story. I really appreciate it. It's an honor. Thank you so much. I love the rapid fire approach. Mike, thanks very much. Keep it right there, everybody. We'll be right back with our next guest live from EMC World. This is The Cube.