Showing posts with label server virtualization. Show all posts
Showing posts with label server virtualization. Show all posts

Tuesday, August 30, 2011

VMworld Case Study: City of Pittsburgh's IT Success and the Beneficial Synergy Between Virtualized Servers and Desktops

Transcript of a BriefingsDirect Podcast from the VMworld 2011 Conference on how the City of Pittsburgh has launched a major virtualization effort to serve a variety of city departments.

Listen to the podcast. Find it on iTunes/iPod and Podcast.com. Download the transcript. Sponsor: VMware.

Dana Gardner: Hello, and welcome to a special BriefingsDirect podcast series coming to you from the VMworld 2011 Conference in Las Vegas. We're here in the week of August 29 to explore the latest in cloud computing and virtualization infrastructure developments.

I'm Dana Gardner, Principal Analyst at Interarbor Solutions, and I’ll be your host throughout this series of VMware-sponsored BriefingsDirect discussions. [Disclosure: VMware is a sponsor of BriefingsDirect podcasts.]

Our next VMworld case study interview focuses on the City of Pittsburgh’s Information Systems organization and how they’ve deeply embraced virtualization at the server level and now increasingly at the desktop level. We’ll see how critical city services in Pittsburgh are being supported using VMware View 4.6 and the new 5.0 version and how the beneficial synergy between virtualized servers and desktops is shaping up.

To learn more, please join me now in welcoming Alex Musicante, the System Security Architect in the City Information Systems department in Pittsburgh. Welcome to the show, Alex.

Alex Musicante: Welcome.

Gardner: Your environment is almost 100 percent virtualized on the server side. First, why is there such a holistic embrace, and how has that provided the confidence for you to move now aggressively into the desktop virtualization space as well?

Musicante: The City of Pittsburgh decided to embrace virtualization five years or so ago, and we did this in a development environment with VMware. The confidence was not there for the server virtualization, and we decided it's a good place to offer development to our internal engineers.

From there, we kept building and building, and we decided to put our first production system on there. Without a problem, everything started going. What virtualization had to offer for us was higher availability, higher reliability.

When we were remote, we had full console access. We were able to offer higher reliability on our development than our production. That was what led us to go to production. It's very difficult in this day and age with budgets and all that. We're now doing more with less. In order to be able to accommodate that and be able to handle the increased workload with fewer people, it has been embracing server virtualization, and virtualization in general.

Gardner: And to what level are you at server-side virtualization and how many servers are involved at this point?

Musicante: In server virtualization we currently have 16 hosts, 98 percent virtual. There are about 250 or so virtual machines (VMs) between two data centers; and we are using VMware Site Recovery Manager to replicate or to bring up the replicated site in the event of a disaster or any planned maintenance that we need to perform at one data center versus the other.

Gardner: I’d like to hear more about your desktop virtualization strategy, but let's learn a little bit more about the scope and scale of your mission-critical set of services. How many users? You're supporting the mayor’s office, police, and fire. Tell us a little bit about what your department is doing in the City of Pittsburgh?

3,000 users

Musicante: The City of Pittsburgh’s City Information Systems Department, which I work for, has about 3,000 users that they support. That ranges from all public safety -- Police, Fire, EMS, and Building Inspection -- to the branches of government -- the Mayor’s Office, the City Council, and Controller’s Office as well as other important departments like the Finance Department, Personnel, Human Resources, and Parks and Recreation. That's who we're supporting, and each and every one of them has their own little caveats of technology that they need.

Gardner: You’re also of course concerned about security, performance, disaster recovery, which you’ve already mentioned. How has virtualization helped you not just in cutting cost, but in making these more hardened, more resilient services?

Musicante: In terms of hardening and security, when we took our virtualization approach, we started out by saying that we were going to physical-to-virtual (P2V) and migrate a lot of these machines. As we proceeded and matured in that environment, we decided that we were going to build fresh and build new.

So when we did our server virtualization, we looked at virtualization in general. It became an opportunity for us to evaluate how we were going to harden things, how we were going to secure things, and since now we don’t have to support that many physical servers, we can expand on our current capacity, and hardware.

We’re able to separate things, where servers that were multi-functional servers, database server, file server, web server, all in one, now get to be three different servers, and only allow communications to the specific application and supports what they need.

Storage came about and offered a lot more flexibility and a lot of benefit to the City of Pittsburgh, but it was not without hassle.



Gardner: Any issues around storage? Has that been something that you’ve been able to wreak some efficiencies out as well?

Musicante: Storage was very interesting for the City of Pittsburgh. They were coming from an environment where everything was on direct-attached storage (DAS), and going to a storage area network (SAN) environment, which they had. They had an array with an HP 6000, but they were only using 500 gigabytes at the time. So storage transition was huge in terms of reliability, but as well as cost at the same time.

It was an unexpected thing from the city’s perspective, as they were not in the market for an array where everything is central. It was all individual and unique to each host and physical server. So storage came about and offered a lot more flexibility and a lot of benefit to the City of Pittsburgh, but it was not without hassle.

Gardner: So you’ve gone through that process -- 98 percent is very impressive on your server, and your infrastructure. What prompted you to now take the additional step to use View and move into desktop virtualization?

Musicante: The City of Pittsburgh moved into desktop virtualization with very similar characteristics as we looked at the server virtualization as how can we offer higher reliability and higher support, give us more management from a central standpoint back at our remote offices, and offer them to the clients and given them the same if not a better level service for additional benefits from administrative.

Security provisioning

There were a bunch of reasons, and those are like pushing out software updates without downtime for the users. They just log off and get a new one. It was security provisioning software, keeping all the storage and everything is back in our data center, so nothing leaves the facility.

Those were motivating factors as well as keeping administrative cost down. That was the push, and it actually took off. It took some time, but it's being embraced more than I ever would have thought it would have been.

Gardner: Let's learn a little bit more about the nature of your distribution requirements. Obviously, you've got City Hall. You’ve got some centralization. You’ve got police headquarters and fire headquarters, but you’ve also got a lot of distributed sites around the city. So let us better understand your distribution requirements when you’re going to desktop virtualization?

Musicante: There are 175 remote facilities, and they range from connectivity of facilities that are on dark fiber, with 100, 200, 300, 500 users, to these individual remote offices that are located in the park facility, and they have one or two employees that are coming across the DSL line.

One of the major complaints was the problem with connectivity where people are on DSL. They would load the roaming profile or pull documents or upload files and they would see this huge lag where it took them upwards of 30 minutes to start their day off. They're now able to go into View, sign-in, and they're in. So we pretty much recovered 30 extra minutes for some of these employees on a daily basis.

Currently, we're in a mixed mode. We have two environments which we're trying to expedite to move off of.



Gardner: How are you leveraging the PCoIP bandwidth improvements for the WAN?

Musicante: Very well. With each version it's definitely gotten better. Still from a management side we do maintain an IPSec tunnel to all of our facilities.

So PC-over-IP has been what we’ve been using for our remote facilities, even back in the 3.0 days. When 4.1 PC-over-IP came out, 4.5, 4.6, it's been progressively getting better and has higher availability with more response. When 4.6, matured, they gave us the View Security Server, and even now with 5, it has increased and lowered the actual requirements necessary for traffic. So some of our facilities are not feeling the same same pain that they were prior to.

Gardner: As you’ve been making this transition, it would be good to understand better how you’ve adopted version 5. To what degree are you using version 5 for View on your desktop virtualization installations?

Musicante: Currently, we're in a mixed mode. We have two environments which we're trying to expedite to move off of, but we currently have a 4.6 environment and a 5.0 environment. Right now with our 5.0 environment we are embracing Persona Management for some of our EMS employees.

Gardner: That’s another one of those ancillary benefits that people don’t always appreciate but it’s pretty important.

Everything is identical

Musicante: Absolutely. It wasn’t something that we were expecting, but at the same time, when we go back with 20/20 hindsight, we reevaluated and said that that makes sense. Everything is now identical. We use non-persistent machine. So every time they log in, it's a brand-new machine and it’s configured identically the way we want it. The only factor that’s different for each user is their profile.

Gardner: You know how to resolve them, it’s not starting from scratch.

Musicante: Absolutely not starting from scratch. That’s also one of the beautiful benefits. As we move and as we mature with the product and as the product matures itself, we seem to be taking a very parallel progression between the two -- the City of Pittsburgh and VMware View. Persona Management right now has been doing wonders for that.

Those departments that have migrated over and wanted to take this “experiment” of Persona Management have been pleasantly surprised. Definitely, that’s also a point to bring up. When you hear problems from people, when end-users complain, there’s always something that they target. It was networking at one point. Then it moved on to virtualization and everyone said it was the promised virtualization, whether it was or wasn’t.

With View, it actually stands alone. It an outlier. Our users call and they say, "I would like to be on View. I would like to be on that system." For an end-user come back to us and request that blows our mind. We appreciate it. It means we’ve done something right. And it also has to be attributed back to VMware. They’ve done something right.

Gardner: Now that you’ve gotten your feet wet, and then some, with 5.0, what are some of the other salient benefits?

So every time they log in, it's a brand-new machine and it’s configured identically the way we want it.



Musicante: That’s going to give us extra 5 percent. There is always that server virtualization where you’d only get that 95 percent, although we got past that. There’s that 5 percent that you couldn’t for or you wouldn’t for whatever reason. That’s the same market for the desktop virtualization and 5 percent was for high graphic intensive people. We're able to now start to achieve that and we're looking to try to achieve that.

We've not gone through some of the advanced 3D accelerated graphic things that are now out with 5. We are in the process of testing, but it’s currently in our test labs within our department. It’s also in terms of deriving the benefit. We have all of our infrastructure. We're going to with a more green approach. So we're going with zero client. They're currently Dell FX 100s. So they may take one tenth of the power, but there is very little there.

I know that VMware View 5.0 3D acceleration is going to be there and is going to help out, but those people are going to be using the repurposed machines, taking their machine, putting a stripped down version of 7 and use it from there. So we're trying to achieve that, but it’s multiple facets.

Gardner: When we think about your adoption pattern around virtualization, you took your time, learned through your development environment, walked in, made some progress and then really ramped up on adoption for your server side. You’ve followed a similar pattern now with desktops.

What’s next? Is there an additional synergy between a private cloud implementation, where you can get even better synergy efficiency? Tell me what you think about this fear and moving towards even higher plane of efficiency and productivity on that overall delivery from a central data center environment?

Going toward the cloud

Musicante: It’s really unclear where we're going to go. As far as cloud and where the cloud is taking the City of Pittsburgh and where the City of Pittsburgh is going with cloud, City of Pittsburgh currently is in the process of taking that last two percent of our system that isn’t virtualized, which is Exchange, and we are currently in the process of going towards the cloud. So it’s actually going to be going to Google Apps for government for mail.

As far as cloud within ourselves, the City of Pittsburgh is using its resources that we’ve regained or recouped from all of our consolidation purposes, especially with the government processes and mentality of doing more with less. There is a lot of fellow government agencies that we're now going to be partnering with to provide them infrastructure as a service.

That’s where some of the other product lines come in like vCloud Director, to be able to allow them to still manage their infrastructure to use our resources, and we can now ourselves be a cloud provider, which I have been marketing as Cloud9 because there are nine entities including the City of Pittsburgh -- nine entities that we are going to consolidate.

Gardner: I'm impressed with the fact that you’ve been able to move through this progression, recoup those savings, and then apply it to the innovation that get you yet more productivity and savings that you can further apply. That’s commendable. Any words of advice for folks that are perhaps not as far along as you’ve been on this progression? What 20/20 hindsight and words of wisdom might you supply them?

Musicante: With server virtualization, everyone is involved in it, and that is the easy part. Desktop virtualization, is where we got hit hard and the lessons that will be learned is that end-user’s matter. Every step of the way, you need their input. It’s not just an administrative decision saying this is the right thing. You need to be good at psychology to convince your users that this is what they want, and getting them to the point of seeing that this is the best approach or getting their input.

The only thing that I could say is to involve your users. Get them in the proof of concept from the beginning.



That really makes all the difference in the world. You’ll have the same end result and you’ll get to the same target, to the same place, but you need their input. It was not the same with server virtualization. That was for the administrators. They owned it. It was their territory. These desktops that you're taking from the users, yes, they’ll have a better reliability, better up-time, better everything, and better end-user experience, but they feel that that’s theirs, and rightly so.

The only thing that I could say is to involve your users. Get them in the proof of concept from the beginning. Get their input, what they need, what they want, how they want to access it, and with that it’ll no doubt be a sure success.

Gardner: And if you can do it in such a way that they are asking by brand for the virtualized desktop approach, that sounds like an optimal outcome.

Musicante: Absolutely.

Gardner: We’ve been talking about the beneficial synergy between virtualized servers and desktops and how that’s shaping up in Pittsburgh. Join me in thanking our guest. We’ve been here with Alex Musicante, System Security Architect in the City Information Systems Department there in Pittsburgh. Thank you so much, Alex.

Musicante: Thank you so much. Have a good one.

Gardner: Thanks, and also thanks to our audience for joining this special podcast coming to you from the 2011 VMworld Conference in Las Vegas.

I'm Dana Gardner, Principal Analyst at Interarbor Solutions, your host throughout this series of VMware-sponsored BriefingsDirect discussions. Thanks again for listening, and come back next time.

Listen to the podcast. Find it on iTunes/iPod and Podcast.com. Download the transcript. Sponsor: VMware.

Transcript of a BriefingsDirect Podcast from the VMworld 2011 Conference on how the City of Pittsburgh has launched a major virtualization effort to serve a variety of city departments. Copyright Interarbor Solutions, LLC, 2005-2011. All rights reserved.

You may also be interested in:

Monday, February 01, 2010

Technology, Process and People Must Combine Smoothly to Achieve Strategic Virtualization Benefits

Transcript of a BriefingsDirect podcast on how to take proper planning, training and management steps to avoid virtualization sprawl and achieve strategic-level benefits.

For more information on virtualization and how it provides a foundation for private clouds, plan to attend the HP Cloud Virtual Conference in March. Register now for this event:
Asia, Pacific, Japan - March 2
Europe Middle East and Africa - March 3
Americas - March 4

Listen to the podcast. Find it on iTunes/iPod and Podcast.com. Download the transcript. Learn more. Sponsor: Hewlett-Packard.

Dana Gardner: Hi, this is Dana Gardner, principal analyst at Interarbor Solutions, and you’re listening to BriefingsDirect. Today, we present a sponsored podcast discussion on planning and implementing data-center virtualization at the strategic-level in enterprises.

Because companies generally begin their use of server virtualization at a tactical level, there is often a complex hurdle in expanding the use of virtualization. Analysts predict that virtualization will support upwards of half of server workloads in just a few years. Yet, we are already seeing gaps between an enterprise’s expectations and their ability to aggressively adopt virtualization without stumbling in some way.

These gaps can involve issues around people, process and technology and often, all three in some combination. Process refinement, proper methodological involvement, and swift problem management often provide proven risk reduction, and provide surefire ways of avoiding pitfalls as virtualization use moves to higher scale.

The goal becomes one of a lifecycle orchestration and governed management approach to virtualization efforts so that the business outcomes, as well as the desired IT efficiencies, are accomplished.

Areas that typically need to be part of any strategic virtualization drive include sufficient education, skilled acquisition, and training. Outsourcing, managed mixed sourcing, and consulting around implementation and operational management are also essential. Then, there are the usual needs around hardware, platforms and system as well as software, testing and integration.

So, we’re here with a panel of Hewlett Packard (HP) executives to examine in-depth the challenges of large scale successful virtualization adoption. We’ll look at how a supplier like HP can help fill the gaps that can hinder virtualization payoffs.

Please join me in welcoming our panel: Tom Clement, worldwide portfolio manager in HP Education Services. Welcome to BriefingsDirect, Tom.

Tom Clement: Thank you, Dana. Great to be here.

Gardner: We're also here with Bob Meyer, virtualization solutions lead with HP Enterprise Business. Hey, Bob.

Bob Meyer: Hey, Dana.

Gardner: And we’re here with Dionne Morgan, worldwide marketing manager at HP Technology Services. Hello, Dionne.

Dionne Morgan: Hello, Dana.

Gardner: Ortega Pittman, worldwide product marketing, HP Enterprise Services, joins us. Hello, Ortega.

Ortega Pittman: Hi, Dana.

Gardner: And lastly, Ryan Reed, worldwide marketing manager at HP Enterprise Business. Hello, Ryan.

Ryan Reed: Hi, Dana, thanks for having me.

Gardner: I want to start by looking at this notion of a doubling of the workload supported by virtualization in just a few years. Why don’t we start with Bob Meyer? Bob, tell me why companies are aggressively approaching the move from islands of servers to now oceans of servers.

Headlong into virtualization

Meyer: Yeah, it's interesting. People, had they known an economic downturn was coming, might have thought that it would have slowed down like the rest of IT spending, but the downturn really forced anybody who is on the front to go headlong into virtualization. Today, we are technically ahead of where we were a year or two years ago with virtualization experience.

Everybody has experience with it. Everybody has significant amounts of virtualization in the production environment. They’ve been able to get a handle on what it can do to see what the real results and tangible benefits are. They can see, especially on the capital expenditure side, what it could do for the budgets and what benefits it can deliver.

Now, looking forward, people realize the benefits, and they are not looking in it just as an endpoint. They're looking down the road and saying, "Okay, this technology is foundational for cloud computing and some other things." Rather than slowing down, we’ll see those workloads increase.

They went from just single percentage points a year and a half ago to 12-15 percent now. Within two years, people are saying it should be about 50 percent. The technology has matured. People have a lot of experience with it. They like what they see in results, and, rather than slow down, it's bringing efficiency to things like the new service model.

Gardner: Ortega Pittman, do you see any other issues around these predictions? The expansion of virtualization seems to be outstripping the skill sets that are available to support it.

Pittman: That's where HP Enterprise Services comes to add value with meeting customers' needs around skills. Many, times small, medium, and large organizations have the needs, but might not have the skills on hand. In providing our outsourcing services, we have the experienced professionals who can step right in and immediately begin the work and the strategic path towards their business outcomes.

The skill demand and the instant ability to get started is something that we take a lot of pride in, and in the global track record of doing that very well is something that HP Enterprise Services can bring from an outsourcing perspective.

Gardner: Dionne Morgan, what are some of the risks, if folks start embarking on this without necessarily thinking it through at a life-cycle level? Are there some examples that you have experienced, where the hope for benefits -- economic and otherwise: agility benefits, flexibility, and elasticity -- somehow end up being imperiled by not being prepared?

Morgan: Many people have probably heard the term "virtual machine sprawl" or "VM sprawl," and that's one of the risks. Part of the reason VM sprawl occurs is because there are no clear defined processes in place to keep the virtualized environment under control.

Virtualization makes it so easy to deploy a new virtual machine or a new server, that if you don’t have the proper processes in place, you could have more and more of the these virtual machines being deployed and you lose control. You lose track of them.

That's why it's very important for our clients to think about not only how they're going to design and build this virtualization solution, but how they're going to continue to manage it on an on-going basis, so they keep it under control and they prevent that VM sprawl from occurring.

Gardner: We’ve talked about this people, process, and technology mixture that needs to come together well. Tom Clement, from that perspective of education, are there things about virtualization that are dramatically or significantly different than what we might consider traditional IT operations or implementation?

Clement: Certainly, there are. When you talk about people, process, and technology, you hit upon the key elements of virtualization project success. There is no doubt in my mind that HP provides the best-in-class virtualization technology to our clients hands down. But, our 30-plus years of experience in providing customer training has shown, time and time again, that technology investments by themselves don’t ensure success.

The business results that clients want in virtualization won’t be achieved until those three elements you just mentioned -- technology, process and people -- are all addressed and aligned.



The business results that clients want in virtualization won’t be achieved until those three elements you just mentioned -- technology, process and people -- are all addressed and aligned.

That's really where training comes in. Our education team can help address both the people and process parts of the equation. Increasing the technical skills of our customers' people is often one of the most effective ways for them to grow, increase their productivity and boost the success rates of their virtualization initiatives.

In fact, an interesting study just last year from IDC found that 60 percent of the factors leading to the general success in the IT function are attributed to the skills of people involved. In that regard, in addition to a suite of technical training, we also offer training in service management, project management, business analysis, all with an eye to helping customers improve and integrate their virtualization projects to better processes -- just as Dionne was speaking about a moment ago -- and to better process management.

Of course, we have stable and experienced instructors, whose practical, hands-on expertise provides clients with valuable tips and tricks that they can immediately use when back on the job. So, Dana, you hit it right on the head. It's when all three of those components -- people, process, and technology -- are addressed, especially in virtualization situations, that customers will maximize the business results that they get back from their virtualization solutions.

Gardner: We’ve also seen in the field that, as people embark on virtualization and move from the tactical to the strategic, it forces a rethinking of what it is core and what might be tangential or commoditized.

Ryan Reed, are we seeing folks who, as they explore virtualization, start also to explore their sourcing options? What are some of the trends that you're seeing around that?

Seeing a shift

Reed: Thank you for asking that question. We do see a shift in the way that IT organizations have considered what they think would be strategic to their end business function. A lot of that is driven through the analysis that goes into planning for a virtual server environment.

When doing something like a virtual server environment, the IT organizations have to take a step back and analyze whether or not this is something that they’ve got the core competency to support. Often times, they come to the conclusion that they don’t have the right set of skills, resources, or locations to support those virtual servers in terms of their data-center location, as well as where those resources are sitting.

So, during the planning of virtual server environments, IT organizations will choose to outsource the planning, the implementation, and the ongoing management of that IT infrastructure to companies like HP.

They apply our best practices and our standard offerings that are available to IT organizations from HP data centers or from data centers that are owned by our clients, which would be considered an on-premise type of virtual server environment. Then, they're managed by the IT professionals that Ortega Pittman had mentioned earlier in either an on-shoring or off-shoring scenario, whichever is the best-case scenario for the IT organization that's looking for that skilled expertise.

It's definitely a good opportunity for IT organizations to take a step back and look at how they want to have that IT infrastructure managed, and often times outsourcing is a part of that conversation.

Gardner: It also sounds like that rethinking allows them to focus on the things that are most important to them, their applications, their business logic, and their business processes and look to someone else to handle the plumbing. In the analogy of a utility, somebody else provides electricity, while they build and manage the motors. Is that fair?

Reed: That's a very fair statement. By choosing a partner to team up with to manage that internal plumbing, as you’d referred to it, it allows the IT organization to get back to basics, to understand how to best provide the best-in-class, lowest-cost service to their end users -- increasing business productivity and helping them maximize the return on their IT investment. This powers the business outcomes that their end-users are looking for.

Gardner: I'm intrigued by this notion that these organizations are going to be encountering virtualization sprawl and trying to expand the use of it, but in different ways are they going to be exercising strengths and weaknesses. What are some of the gaps that are typical? What do we usually see now in the field that create a stumbling block to the wider adoption of virtualization?

Pittman: One of the things we observe in the industry is that many customers will start with a kind of phase one of virtualization. They'll consolidate their servers and maybe stop just there. They get that front-end benefit, but that exhausts the internal plumbing that you referred to in a lot of different ways, and can actually cause challenges and complexities that were not in their immediate expectation. So, it's a challenge to think that you're going to start with virtualization and not go beyond the hypervisor.

The starting point

We’d like to work with our customers to understand that it's the starting point to consolidate, but there is a lot more in the broader ecosystem consider, as they think about optimizing their environment.

One of HP’s philosophies is the whole concept of converged infrastructure. That's thinking about the infrastructure more holistically and addressing the applications, as you said, as well as your server environments and not doing one off, but looking more holistically to get the full benefit.

Moving forward, that's something that we certainly could help customers do from an outsourcing standpoint in enabling all of the parts, so there aren’t gaps that cause bigger problems than the one hiccup that started the whole notion of virtualization in the beginning.

Gardner: Does anyone else has some observations from the field about what gaps these organizations are encountering as they try to expand virtualization use?

Clement: One of the good things for our clients is the fact that within HP we have a great deal of experience and knowledge regarding virtualization. Through no fault of their own, many clients don’t understand or don’t realize the breadth or depth of virtualization options and alternatives that are available for them.

We want to make sure that the customers are thinking about this first from the business perspective.



The good news is that we at HP have a wide range of training services, ways that we can work with a client to help them figure out what the best implementation options are for them, and then for us to help them make sure that those options are implemented with excellence and truly do result in the business benefits that they desire.

Gardner: Now that you’ve mentioned some of the strengths that HP is bringing to the table, how do you get those to work in concert? It seems that it's a hurdle for these organizations themselves to look at things holistically? When they go out to a supplier that has so many different strengths and offerings, how do you customize those offerings individually to these organizations. How do they get started?

Morgan: We think about this in terms of their life cycle. We like to start with a strategy discussion, where we have consultants sit down with the client to better understand what they’re trying to accomplish from a business objective perspective. We want to make sure that the customers are thinking about this first from the business perspective. What are their goals? What are they trying to accomplish? And, how can virtualization help them accomplish those goals?

Then, we also can help them with their actual return on investment (ROI) analysis and we have ROI tools that we can use to help them develop that analysis. We have experts to help them with the business justification. We try to take it from a business approach first and then design the right virtualization solution to help them accomplish those goals.

Gardner: It sounds like there's a management element here. As we pointed out a little earlier, IT departments themselves have been divvied up by the type of infrastructure that they were responsible for. That certainly makes a lot of sense, and it follows the development of these different technologies at different times in the past.

Now, we're asking them, as we virtualize, to take an entirely different look, which is more horizontal across this converged infrastructure. Is there a management gap that needs to be filled or at least acknowledged and adjusted to in terms of how IT departments run?

Blurring the connections

Meyer: What it calls into focus is that one thing virtualization does very nicely is blur the connections between the various pieces of infrastructure, and the technology has developed quite a bit to allow that to ebb and flow with the business needs.

And, you're right. The other side of that is getting the people to actually work and plan together. We always talk about virtualization as not an end-point. It's an enabler of technology to get you there.

If you put what we’re talking about in context, the next thing that people want to go to is maybe build a private-cloud service delivery model. Those types of things will depend on that cooperation. It's not just virtualization that that's causing but it's really the newest service delivery models. Where people are heading with their services absolutely requires management and a look at new processes as well.

Gardner: In many cases, that requires a third party of some sort to be involved, at least, to get that management shift or acknowledgment under way.

Which of you can offer an example of how we move to a higher level of virtualization and got those payoffs that people are so enticed by -- that much lower number of servers, lower footprint, lower carbon and energy use, total cost, etc.? Can you provide an example of an organization that's done that and has also bitten the bullet on some of the management issues that allows that economic benefit?

They decided to virtualize, because that would help, of course, with the ability to consolidate and to improve on those service levels.



Morgan: I can give one example. There's an organization called Intrum Justitia, a financial services organization in Europe. We worked with them as they were embarking out their virtualization journey. The challenge they had was that they have multiple organizations and multiple data centers across Europe, and they wanted to consolidate from 40 different locations around Europe into two data centers.

At the same time, they wanted to improve the service level they were providing back to their business. They decided to virtualize, because that would help, of course, with the ability to consolidate and to improve on those service levels.

The way we helped them was by first having that strategy discussion. Then, we helped them design the solution, which included the HP Blade System, VMware software, EVA Storage, as well as other hardware and software products. We went through the full lifecycle with them helping with the strategy and the design.

We helped them build the solution. We managed their project office. We managed the migration from the 40 locations. Then, once everything was transitioned, we were able to help them go on the right path to further managing them. Some of the results were that they were able to manage that consolidation to the twin data centers, and they're beginning to see some of the benefits now.

Gardner: Let me put you on the line. What do you think HP brought to the table in this example that the Intrum wouldn’t be able to find anywhere else?

For more information on HP's Virtual Services, please go to: www.hp.com/go/virtualization and www.hp.com/go/services.

Wide expertise

Morgan: There are a couple of things. One is that we actually have the expertise, not only in the HP products, but also in the software products. We have the expertise, of course, for the Blade Systems and the EVA Storage, but also the expertise around VMware.

So, they had hardware and software expertise from one vendor -- from HP. We also have the expertise across the lifecycle, so they could just come to one place for strategy, design, development, and the ultimate migration and implementation. It's expertise, as well as a comprehensive focused life goal.

Gardner: Are there any other examples of a larger scale, top tier organization that has moved aggressively into virtualization and had a success?

Pittman: Yes, Dana, HP Enterprise Services worked with the Navy/Marine Corps Intranet (NMCI), which is the world’s largest private network, serving and supporting sailors, marines, and civilians in more than 620 locations worldwide.

They were experiencing business challenges in productivity and innovation and in the security areas. Our approach was to consolidate 2,700 physical servers down to 300, reducing outage minutes by almost half. This decreased NMCI’s IT footprint by almost 40 percent and cut carbon emissions by almost 7,000 tons.

We minimized their downtime and controlled cost. We accelerated transfer times, transparency and optimal performance.



Virtualizing the servers in this environment enabled them to eliminate carbon emissions equivalent to taking 3,600 cars off the road for one year. So, there were tremendous improvements in that area. We minimized their downtime and controlled cost. We accelerated transfer times, transparency and optimal performance.

All of this was done through the outsourcing virtualization support of HP Enterprise Services and we're really proud that that had a huge impact. They were recognized for an award, as a result of this virtualization improvement, which was pretty outstanding. We talked a little earlier about the broader benefits that customers can expect, the services that help make all of this happen.

In our full portfolio within the IT organization of HP, that would be server management services, data center modernization, network application services, storage services, web hosting services, and network management services. All combined, they made this happen successfully. We're really proud of that, and that's an example of the very large-scale impact that's reaping a lot of benefit.

Gardner: We've talked about how this can scale up, I suppose it's also interesting in the future, as more companies look to virtualization and think about services and infrastructure as a service (IaaS), that this could probably start going down market as well. Does anyone have some thoughts about how a company like HP, perhaps through their outsourcing capabilities, could move somebody’s values into an organization smaller than the Navy and Marines?

Mission-critical systems

Reed: What's interesting about the NMCI is that, as Ortega mentioned, this is a very large complex and mission-critical system. Thousands of servers were virtualized, having a major impact on how the service is being delivered. The missions that are being performed on such an infrastructure are still mission critical. You can't really have a much more impactful implication, because lives actually depend on the successful missions that are performed on this infrastructure.

Now, if you take that and have it scaled to lower level implementations of virtual server environments, the lessons learned, the best practices, the technology, the people, the processes, and the skills are all absolutely relevant, when trying to scale this down to small- and medium-sized businesses.

That's because the standardized procedures for managing this type of infrastructure is documented for our service delivery organizations around the world to take advantage of. They’re repeatable, standardized, and consistently delivered.

Gardner: As we get into the future, and the use of virtualization becomes integral to more companies -- not as an island, but more of the ocean that they are sailing on -- this kind of changes the way the companies function. They'll become more IT services and service management oriented. Perhaps, they'll have more services orientation in terms of their architecture.

Does anyone have any thoughts about where this is going to lead next, if you bite the bullet, become holistically adept at virtualization partnering with companies like HP to use the skills and understanding they have and learn the lessons of the past? What are the next stages or steps? Bob, any thoughts?

Virtualization becomes a foundational element for the next set of service delivery model that people are looking at.



Meyer: We mentioned this in the beginning. Virtualization becomes a foundational element for the next set of service delivery model that people are looking at. So, from an IT provider’s perspective, if you get virtualization right, if you get the converged infrastructure that Ortega was talking about, you get the right mix and close the skill gaps. You get a strong foundation to move on to things like private cloud, and it really opens up your options for different service delivery models.

With this is this notion of pushing out virtualization more broadly, the next step leads you to a good place to build on top of those delivery models and ultimately lower the cost and increase the quality of the services you deliver to the business.

Pittman: You asked how it all fits in moving to the future. Recently, in a Gartner report, there were some key findings. One of the items that was reported was that mid-sized businesses are seeking a much more intimate relationship with IT providers. There is a perception out there that they can have a closer relationship with smaller vendors as opposed the large ones.

[Editor's Note: “The penetration of virtual machines in the market at year-end 2008 was 12%; by year-end 2012, it will be nearly 50%.” Source: Gartner October 7 2009. Research Title: Virtual Machines and Market Share through 20012. Research ID #G00170437.]

One thing I’d like to just put out there for the IT community that may be is thinking about virtualization is that HP offers solutions for small, medium and large organizations. The way we are set up in terms of the account support with our account leaders, we certainly can meet the needs of the small to medium to large menu. We are set up to engage, support, and be that trusted advisor at all three of those levels.

Just to dispel any misconception that "They’re large, and I'm not sure if I'm going to get the attention," we're ready and have the products and services to deliver outcomes that they are looking for at all levels.

Gardner: Sort of "have it your way" opportunity.

Pittman: Exactly.

Expertise and flexibility

Clement: Just to follow on to that point, which I think is a great one. As we've been hearing here, it boils down to expertise and flexibility. Does HP have the expertise strategically to help clients of any size? Do we have the expertise from a service delivery perspective, from an instructor perspective, from a course development perspective? And the answer is, we do.

Do we provide these services, these products, these training classes in a variety of flexible ways and are we willing to tailor these to our clients. The answer, again, is a resounding, yes, we are.

Gardner: I wonder if we could offer some concrete ways to get started. Are there some places people can go, some Google searches they should do, as they are thinking about virtualization and their expansion and their way of managing the risk?

Morgan: There is definitely HP.com. We have many pages on HP.com to talk about virtualization and our virtualization offerings. So, that is the one area. They could also contact their local HP representative. If they work with HP authorized channel partners, they can also have discussions with the channel partners as well.

Meyer: There's a very simple way to find out more about virtualization solutions. You could just type in www.hp.com/go/virtualization, and it will take you to virtualization home page. If specifically you want to find more about services, it's just www.hp.com/go/services. That shortcut will take you right to the very relevant information.

Gardner: Well, very good. We've been here with a panel of HP executives examining the in-depth challenges of moving to large scale successful virtualization adoption. We looked at some of the ways that HP has worked with some customers to help them make that leap successfully. I want to thank our panel today. We've been talking with Tom Clement, worldwide portfolio manager in HP Education Services. Thank you, Tom.

Clement: You're most welcome, Dana. Again, thanks for having me.

Gardner: Bob Meyer, virtualization solutions lead, HP Enterprise Business. Thank you Bob.

Meyer: Thank you.

Gardner: Dionne Morgan, worldwide marketing manager, HP Technology Services. Thank you, Dionne.

Morgan: You're welcome.

Gardner: Ortega Pittman, worldwide product marketing, HP Enterprise Services. Thank you.

Pittman: Thank you for having me.

Gardner: And, Ryan Reed, worldwide marketing manager, HP Enterprise Services.

This is Dana Gardner, principal analyst at Interarbor Solutions. You've been listening to a sponsored BriefingsDirect podcast. Thanks for listening and come back next time.

Listen to the podcast. Find it on iTunes/iPod and Podcast.com. Download the transcript. Learn more. Sponsor: Hewlett-Packard.

Transcript of a BriefingsDirect podcast on how to take proper planning, training and management steps to avoid virtualization sprawl and achieve strategic-level benefits. Copyright Interarbor Solutions, LLC, 2005-2010. All rights reserved.

For more information on virtualization and how it provides a foundation for private clouds, plan to attend the HP Cloud Virtual Conference in March. Register now for this event:
Asia, Pacific, Japan - March 2
Europe Middle East and Africa - March 3
Americas - March 4

You may also be interested in: