Technologies like virtualization and cloud computing promise enormous leaps in efficiency and flexibility, but they can lead organizations into a quagmire if they don't plan properly for the transition, says Bill Hurley, CIO, CTO and executive vice president of Westcon Group. Without proper planning, organizations can stall in the midst of their transitions to virtualized environments or the cloud, finding themselves with a bundle of sunk costs and no path forward.
Westcon Group is a value-added distributor of unified communications, network infrastructure, data center and security solutions that has made its own transition to a 100 percent virtualized environment and now helps its reseller, system integrator and service provider customers guide their own customers through the transition to virtualized environments and the cloud.
The Path to Virtualization
"About three years ago, we realized we needed to consolidate our data centers," Hurley says, explaining that the firm--which has about 2,300 employees in 60 offices in more than 30 countries--had data centers in New York and London at the time. The original idea behind the two data centers was that they would back each other up. But after years of maintenance by different teams, the two data centers had different configurations and architectures. They were no longer capable of the task and the technology was getting long in the tooth to boot.
"We came to a quick realization that we could do better with a single data center with a hot site for backup," Hurley says. "About one-third of our platform needed to be refreshed."
But while Hurley's organization was planning the consolidation, a number of engineers in the IT department had started a skunkworks project of their own. Hurley explains the engineers were maintaining servers that were crashing frequently, and they wanted to sleep at night. So they started an initiative to virtualize some of the older, more crash-prone servers. The skunkworks project led to real improvements, so Westcon decided to formalize it, add some money and turn it into a real project.
"We were just going to replace our data center with a like configuration." Hurley says. "But then we started to realize this virtualization thing could really save us a lot of money. We could virtualize and greatly reduce our footprint. Instead of spending $1 million to refresh one-third of our technology, we could replace the entire platform for roughly that $1 million."
Working with Cisco, it made the transition. It first virtualized its environment as it was in the existing data centers, and then moved the virtual machines (VMs) to its new data center. Hurley says the result was a two-thirds reduction in heating and cooling costs, a reduction in the number of engineers focused on supporting equipment and applying patches and an overall more agile environment. He also notes the move greatly reduced Westcon's outages.
"We received a tremendous number of benefits over and above our initial ROI," he says.
Westcon successfully managed its transition. But many organizations fail or stall in their move to virtualization or to the cloud, Hurley says.
"Virtualization stall, when we see it, happens primarily at the enterprise level," Hurley says. These projects often start out like Westcon's own skunkworks project, but fail to get off the ground because IT is either unable to justify the ROI or business value of the project or they underestimate the total cost of the project.
"They end up judging the ROI based on speeds and feeds rather than the business value offered to the enterprise," Hurley says. "It's got to be about the business initiatives running on those pieces of equipment you're about to virtualize. They don't get the budgets approved because they're not able to talk to the business in terms of the ROI value of the project."
While failure to explain the business value of the project can torpedo it before it gets funding, underestimating the actual cost of virtualization is another leading factor in virtualization stall, Hurley says. Organizations often base their estimates solely on the cost of the VMs and fail to consider the concomitant infrastructure costs for storage, networking and security.
"For every dollar spent on virtualization software, the IT manager should expect to spend another $11 or $12 on infrastructure for that virtualization project," Hurley says. "They end up running out of money before they achieve the effects they were looking for."
Hurley calls virtualization the gasoline that enables the cloud. More and more organizations are investigating the benefits of cloud computing and beginning deployments. But cloud computing has its pitfalls. Avoiding cloud stall also requires proper planning, Hurley says, because the way you use software inside the enterprise may not be the same way it will be used in the cloud.
"There's a tremendous amount of preparation work that needs to be done before you jump into the cloud," he says. "You need business process analysis to make sure I don't just take my Exchange servers, push them into the cloud, and then find out I have someone in France that's running a 200,000 item inbox in their exchange server that grinds everything to a halt. Everybody's all hot and bothered by $5 per person per month for email, but you really need to spend an extra $50,000 or $100,000 up front to make sure you're not going to be surprised when you push things into the cloud."
Some firms also assume that moving to the cloud means they can reduce or eliminate their IT function, and that's a huge mistake that can lead to stall, Hurley says.
"I'm always amazed that the commentary with the cloud is that you don't need IT anymore," he says, noting that cloud services often change the focus of IT, but they don't eliminate the need. Organizations still need to keep their networks up and running, their security working and they will need to integrate their new cloud services with business processes to realize its value.
"That integration effort can be complicated, depending on the maturity of the cloud service provider and the offering they have," Hurley says.
Another potential chokepoint, and one frequently overlooked, is the depreciation schedule for servers and other hardware.
"We're spending a lot of time helping the channel understand the financial implications of going to the cloud," Hurley says. "If the IT department the year before spent $500,000 for servers and other hardware to run your CRM, if you decide not to use that hardware anymore, the CFO is going to have to accelerate that write off."
For some firms it will make financial sense to go forward to the cloud despite these issues, Hurley says, and other firms that run the numbers will find that it is better to wait and leverage their existing technology for a few years before taking the leap. In any case, the best practice is to analyze the costs and benefits before taking action.
"In that ROI analysis, you might be pointing out all the savings, but you're also going to have to buy bigger pipes for the bandwidth to the Internet," Hurley says. "And what about all that hardware and software the company just bought for you last year? It may be that the ROI still works. But make sure. The stall factor happens when there are surprises."
Thor Olavsrud is a senior writer for CIO.com. Follow him @ThorOlavsrud.
Read more about project management in CIO's Project Management Drilldown.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.