 Extracting the signal from the noise, it's theCUBE, covering VMworld 2015. Brought to you by VMware and its ecosystem sponsors. And now your host, Dave Vellante. We're back, welcome to theCUBE. We're here at Moscone North. This is VMworld 2015 and this is theCUBE. theCUBE goes out, we go to all our events, we extract the signal from the noise and if you want signal, you got to talk to IT practitioners. So we love that practitioner view. We have peers talking to peers at this event. 23,000 people and there's a lot of customers that are helping each other solve problems and just getting some new ideas. Jay Benson is here. He's an IT practitioner from eHealth in Saskatchewan, Canada. Jay, welcome to theCUBE. Thanks very much for taking the time to come on. Thank you, great to be here. So we were talking off camera. This is your first VMworld. It's a great event that's really evolved over the year but the core of it is the IT practitioner audience. It's a show for guys like you. So what are your initial impressions? Well, it's an impressive place. 23,000 plus people, it's phenomenal. Just the amount of people and knowledge sharing that's going on, it's learning opportunities. It's absolutely incredible. Any particular highlights for you this year? Things that you were hoping to learn or any surprises or other things that you didn't expect that you're going to take back? Well, you mentioned talking to practitioners. That's part of what I like to do is talk to others that are trying to do similar things to what we're doing and leverage that knowledge and leverage that experience and learn from others is pretty huge for me. And so looking forward to that and have engaged in a lot of that activity already. Talk about e-health. Tell us about the organization and the role there. Sure, so e-health Saskatchewan is a crown corporation within the provinces of Saskatchewan. We look after and are mandated to support the electronic health record for the provinces of Saskatchewan. So we look after essentially the medical history of all of the citizens of the province, just over a million people in terms of things like diagnostic imagery, prescriptions, immunization records, birth and death records, things of that kind in Asia, clinical encounters, so that our medical practitioners have a full history of every person's visit and journey through the health system as they work and as they move. So I interview a lot of healthcare practitioners and providers in the US and the health system here is quite a bit different. Yes, it is. And so the drivers and the pressures are all around, certainly around compliance, I'm sure you've faced many of those, but things like electronic medical records and here it's meaningful use and they demonstrate use in order to get paid. What are the drivers in Canada? It's very similar in terms of providing information and things of that nature, but as you mentioned, healthcare in Canada is government funded. And in days of reduced budgets and doing more with less, that's really the goal for all of us is to make sure that we're delivering an excellent service, a service that our practitioners use because ultimately it's about patient care but doing so in a cost-effective manner, which is one of our core drivers. We're always architecting for cost is maybe the best way to think of that as well as capability and functionality, obviously. Yeah, so those things sometimes are counterposed, but yeah, it makes your job interesting. So, all right, so let's talk about your environment, your infrastructure, the applications, just put a picture for us. Sure, so we're a single data center today and we support, as I mentioned, electronic health records for just over a million people in the province. Each of those systems that I mentioned, pharmacy, electronic records, immunization records, things of that nature are all disparate systems today, which means we do an awful lot of work around integrations. Integrations is core to what we do. There's also within the province, there's 13 individual health regions that look after a lot of their own systems too that we have to collect at a provincial level to aggregate that record for everybody, which means not only are we integrating with ourselves, we're integrating with our partners within the health sector, and that's an awful lot of work, as you can imagine. It's an awful lot of data normalization, data stewardship type issues, things of that nature that are just core required to come up with one single record for every individual person. So obviously you're a VMware account. We are. So, I mean, can you give us a picture? Sure. How many servers, how much storage, number of VMs and all the things. So we're currently sitting at just over a thousand production machines. They're roughly a 60%, 40% split between the VMware and Intel stack and a mid-range stack. So we cover an awful lot of technology within our scenario. On the Intel side of that equation or the VMware side of that equation, we're about 90% virtualized today, 95% virtualized today, and that's just over 800 production servers, almost twice that in terms of our non-production environments between development and user acceptance testing. Okay, so obviously, so you know it's funny, we've been talking all week, we started doing theCUBE in 2010 at VMware, and then the discussion back then was, well, what percent virtualized are you? And it's like, well, I'm 30%, and I want to get to 50% by next year, all right? So you've been through that journey. So now it's like, now it's cloud and it's mobile, and so you've got a whole new set of challenges, and one of the ones that we're going to talk about here is sort of end user computing and mobile. So, and that place is huge pressure on the infrastructure by talking about traffic moving east-west instead of just north-south. Absolutely. You've got flash coming in and moving bottlenecks around and everything just, once you think you get to the end of that journey, boom, a whole new one begins. Yeah, start again. Great, so you're on a new S-curve, love it. So let's talk about sort of the end user computing piece, what you're doing. Some people call it VDI or whatever the new word is. Yeah, so we might move around a couple of things here, but fundamentally, we're working on three major projects right now. End user computing or VDI is one of those. But in order to explain that, I got to back up just a little bit. So we're also in the midst of standing up two new data centers. As I mentioned, we are single today. We're moving to a dual data center model, and the intention there is redundancy and reliability for our end users and our practitioners. So that involves a significant amount of work in terms of new equipment, new hardware, new gear, new setups, all of which is gonna run what is essentially a private cloud for the health sector within the province. And one of the core services that's gonna be offered out of that private cloud is end user computing or VDI. Okay. So there are a series of steps for us. And particularly in the end user compute space, our intention is around delivering a consistent experience for our end users. Because when it comes right down to it, Saskatchewan's a fairly large geography in terms of comparison. We're about the size of Texas, but only just over a million people, which means we're spread out across the entire province, and which means our facilities and the places where our patients go are spread out across that province. But to centralize everything in two data centers, we worry about latency, we worry about user experience. So end user computing is gonna do a couple of things for us. The biggest one is to bring those applications, bring those desktops closer to our applications within the data center, allows the power of the data center to expose that speed to our end users on the slow end of those remote links where all we're passing is screenshots for them. The other thing it does is security. As you can imagine with medical records, we're all about personally identifiable information, we're all about legislation and worrying about keeping control of that. So keeping that data within our data centers allows us that control, and but at the same time allows people to access and use it, so VDI is very key to us. So let's unpack that project a little bit. So you're putting in an end user computing solution from EMC, we had those guys on earlier this week and they were sort of describing that stack. You're in the early stages of doing that, but so what's the problem that you're trying to solve and what do you actually deploy? Sure, so I'm gonna back up a little bit in order to set some more context for you. eHouse Saskatchewan, from an IT perspective, we're also looking at moving towards a managed service provider approach for the health sector. So moving away from that traditional enterprise and more to a service delivery type of model. And one of the things that we're offering as part of that is desktop as a service, because as I mentioned, those 13 health regions that we have tend to do things in disparate ways. It's just simply because they're 13 organizations, each with their own budget, each with their own approach. What happens unfortunately though with that is an inconsistent approach, particularly around user experience, right? A lot of our practitioners tend to move between regions, which means they do things one way in one office, one thing, one way in a different facility, and that causes them angst and concern as you can imagine. So we're looking to standardize that and standardize that delivery, which will make their experience better, right? In order to do that, we had to look for a solution that would help us get down that path very quickly, and that's where EMC and the Federation and VMware came into the fore for us. You've been talking to others all week about the end user compute platform and what's happening there. For us, that was a key differentiator. It was around time to market, it was around an engineered solution that we know will work when we get there, as opposed to we could build it ourselves, we could have built it ourselves, but that's a lengthy endeavor. We figured that was 18 to 24 months minimum. We're not sure we can wait that long when it comes right down to it. The demands on us around that user experience, around some new facilities that are being built and new models of delivery within the province around efficiency and sessions following me from room to room from our practitioner standpoint, that's absolutely key, and we can't wait two years to get to that. So the time to market of this particular piece and the VMware stack that allows the orchestration and the automation and the self-serve and the cost models are absolutely imperative to us. It's just the way it needs to be. So what does that, I recall from the description of the solution, it had some flexibility in there in terms of the storage infrastructure, obviously it's UCS or the VBlock, but then the higher levels up the stack. So what did you actually, how did you compose that solution? Because it was semi-composable. If I could... Yeah, so there are some pieces that you could add or subtract as needed. We've been an EMC customer for many years in terms of our storage approach and I wanted to continue to leverage that investment. So from a storage perspective, we're very much a build our own kind of an approach here. We've made some significant investments under the new data center project around individual chunks of hardware, both compute and storage. And we decided to leave that in place. So one of the core functionality or flexibility pieces that we gained out of the UC solution was the fact that we could almost build our own and then layer the software stack on top of it. So it allows us to continue to leverage the EMC investment we've already made in terms of core storage and data replication and data redundancy pieces, layer the UCS platform on top of that from a Cisco perspective and compute, and then finally layer the VMWare software stack on top of that to get us to that point of higher virtualization, higher portability, consistent user experience and the ability to wrap automation and orchestration around all of that to make our staff and our team much more efficient. And then you're doing NSX as part of this? In a small way, yes. So you're dipping your toe in NSX? We're dipping our toes in NSX, yes. You get, again, let's squint through the market, right? NSX is gonna, damn, it's everywhere. Okay, and you talk to practitioners, they're like, yeah, you know, we're interested and we're trying it. So what's your take on NSX? Sure, we are interested. And I certainly see NSX and by extension software defined networking in our roadmap and in our go forward plans. What we're trying to do now is, as you said, just dip our toes in the water. Part of that, particularly with the end user computing solution, I mentioned two data centers that we're gonna run this out of. NSX in this case is gonna be used to help insist with the high availability of the management nodes that will control that end user experience. We wanna make sure that those nodes are available in either data center when needed and part of the engineered solution relies on NSX to do that. As we move forward, items like policy driven security, like microsegmentation, we see a lot of potential in those and expect at some point in our roadmap we will take advantage of them. Right now we're, as we said, kicking the tires a little bit and trying to see how that will actually apply in a healthcare setting. Make sure we can leverage it with some of our legacy applications that we still need to deal with. Jay, as you bring together these so-called hardware-converged solutions, whether it's hardware-converged solutions or things like NSX, which is one of the software convergence, how does your organization change or is it your sort of smaller organization and you don't have all these sort of silos of pockets of expertise centers? Is you're more looking for that solution? Can you describe that? Sure, sure. So, very much today we're a traditional IT shop as you can imagine, individual people that are there to deal with each silo or tower of technology, storage people, network people. We have expertise centers. We always talk about silos in a pejorative, but there's a good side of silos, too, is you've got expertise, you've got deep guys that really know their stuff. But yeah, they're off in their silos, that's true. And they're very good at it. And we actually have a very good team and I put them against just about anybody. But what we're seeing as part of this convergence is and more software defined is they're gonna become less a silo individual team and more of a data center services or data center operations team. And that's really the view. In fact, all of those operations teams within the organization currently report to one person. And his title is manager data center services. And that's why, is to try to drive some of those synergies so that we can leverage the expertise that these people have in a little more efficient manner. Today, the very traditional workflow of, when I request a new machine, I have to go through the steps to get the servers provisioned and the storage provisioned and the load balancers set up and the applications installed. And when it comes right down to it, our goal is to provide a better service to our medical practitioners within the province which ultimately relates to patient care. And when it comes right down to it, doing those individual steps doesn't really add that value. What I'd much rather those people doing is worrying about the availability, the reliability, the user experience that goes with that, getting the right data to the right people at the right time in the right manner. And we see automation and standardization as a way to do that. Which almost inevitably will lead to the convergence of those towers. You're embracing that automation. Absolutely. A lot of IT practitioners, I talk to, they say, yeah, we're headed in that direction but they're afraid of automation. Because I don't want to give up control. They like having the knobs and dials to turn. But are you feeling like you're through that inertia? We're a little bit early in that one yet. We're still working through some of that. Change management, organizational change management is always tough, right? And there's a lot of different ways of doing things. And you're right, control is a big thing. Most IT people like control. And control freaks for good reasons. Exactly, for very good reason. That may sound negative and it isn't, right? It's a very good point. What we are trying to do is move to more of that standard way of doing things. Which allows us to take all of the best practices that they already know and wrap it into a repeatable process so that it isn't onerous on those people to do the same job anymore. It allows us to slow down the treadmill that they're on a little bit and provide a little different value. So your justification for this EMC, you know, EUC solution was the standardization piece. Yes. Primarily. Yep. And the solution itself was differentiable and that's sort of what was the allure. Did you have to quantify all that? To a certain extent, yes. As you can imagine with any type of large investment like this, you're looking at business cases to do that and everything else. So we did spend a significant amount of time looking at how our organization operates, the services that we can provide to the rest of the health sector. And looking at particularly, again, that user experience, the security pieces and the delivery mechanisms that we have. As a fairly distributed province, we worry about latency, as I mentioned before. We spend a lot of time measuring that latency and looking at the potential impact of centralizing that application delivery out of two data centers. To try to help quantify our approach and why we're moving down this path. You mentioned it's a federation solution. Yes, it is. There's a lot of talk leading up to the show about the whole federation. I'm always interested in the practitioner's view because there's not a lot of talk on the floor about what's going on with EMC and VMware and management and all that stuff. From an IT practitioner standpoint, do you pay attention to that stuff? Do you care? Obviously, you like the federation solution. Is the federation of value to you or is it just this one point case and what are your thoughts on the whole thing? It's absolutely of value to us actually. One of the, we're all about outcomes. We're all about getting to the end of the job and when it comes right down to it, the expectation we have is that this stuff has to work and the way we see to do that is items like the federation pieces where you're taking multiple disparate technologies from multiple disparate companies and you're actually putting wrapping engineering hours around them to ensure that everything works together the way it should and the way it's supposed to and then adding on top of that a single support call which means I can avoid the normal finger pointing that you tend to get into when you have multiple vendors involved in a solution. Well, mine works, well, no, mine works. No, the problem's over there and you just don't get anything done. The fact that this is an agreed upon federation approach already with multiple organizations all driving to that same outcome is of significant benefit to us. It allows us to deal with that single support call. It allows us to have that solution built with our staff and decide them but in a manner that we know is going to work and get a significantly further down the path in a hurry as opposed to again doing it ourselves like I mentioned. Awesome, well, Jay, thanks very much for coming to theCUBE. It was really a pleasure meeting you. You too. Sharing your practitioner knowledge with our audience. Great stuff. Great, thank you very much. You're welcome. All right, keep right there, everybody. We'll be back with our next guest. This is theCUBE, we're live from Moscone at VMworld 2015. Right back.