The End of Windows Server 2003 Means It's Time to Consider the Cloud

For many IT decision-makers, the cloud remains a spectator sport. They have attentively observed the cloud’s emergence as a technology provisioning alternative, and they have read about its adoption by others with interest. But when it comes to their own data centers, they’ve maintained a wait-and-see mentality.

The time for sitting on the sidelines, however, is rapidly coming to an end. Several triggers are compelling decision-makers to modernize their IT environments sooner, rather than later. And, more often than not, that modernization includes the cloud.

The Last Days of Microsoft Windows Server 2003

One reality that is prompting many organizations to plan at least a partial move to the cloud is the end of support for Microsoft Windows Server 2003, which will occur on July 14, 2015. This date is important for two reasons. First, more than 23 million instances of Windows Server 2003 have been deployed to support all kinds of workloads — from Microsoft’s own office productivity tools to custom business applications. The termination of support for Windows Server 2003 will potentially affect a tremendous number of organizations across all market segments.

Second, the end of support for Windows Server 2003 will make its continued use untenable. Without further security updates from Microsoft, the platform will become unacceptably vulnerable to attack. Technical support for an out-of-service platform will likely be impractically expensive. And third-party software vendors will stop supporting it as well.

IT decision-makers must therefore start planning and executing their next data center move right away. More specifically, they must decide if they want to migrate to a more modern operating system such as Windows Server 2012 using their existing infrastructure — or if it’s time to do a more comprehensive data center refresh.

“The end of support for Windows Server 2003 offers IT decision-makers a great opportunity to take a fresh look at how they are provisioning all of the applications and services in their portfolios,” says CDW Cloud Client Executive Tracy David. “Based on that evaluation, they can move forward with a strategic data center modernization plan designed to address both their immediate and long-term workload requirements.”

Keeping Pace

Another factor pushing IT decision-makers to consider the cloud is the fact that some users in their organizations may already there. Many users have embraced Software as a Service (SaaS) offerings that don’t require an IT department’s involvement at all. Often referred to as “shadow IT,” this phenomenon has a variety of consequences. One is that business users are becoming less patient with IT departments that they perceive to be slow to respond to their needs, because they can get what they want quickly and easily elsewhere. Another is that IT managers’ reservations about moving to the cloud are becoming moot — since it is happening anyway.

But left to their own devices, users can make bad decisions. “With the advent of the cloud, a lot of business users think all you need is a credit card and an Internet connection to do IT,” says David. “They don’t understand the consequences of going to the cloud without considering issues such as security, integration, data management and business continuity.”

Organizations also can be in danger of getting hit by runaway costs when users adopt the cloud at their own discretion. “SaaS can make sense in many circumstances,” says Gee Rittenhouse, senior vice president and general manager of Cisco Systems’ Cloud and Virtualization Group. “But it can also become inordinately expensive if you depend on it too much, because you’re paying every vendor redundantly to manage underlying infrastructure and services for your applications and data.”

Rittenhouse says IT departments should exercise oversight over cloud adoption and, where appropriate, run applications on a private-cloud infrastructure. “You can definitely reduce IT costs with a well-automated, highly utilized private cloud that’s amortized across all of your workloads,” he says.

The Quest for Performance-on-Demand

A third factor driving cloud adoption is the need for greater capacity and performance. Organizations are generating more data internally and tapping into more data from external sources. To get maximum value from this data, they are adopting sophisticated analytic and reporting tools. They are also empowering employees, customers, suppliers and partners to make use of this data — and to collaborate in other ways — on an anywhere, anytime basis using their desktops, smartphones, tablets and other devices.

The result is more intense computing, storage and network workloads that out-of-date data center infrastructure simply isn’t equipped to handle. Many of these demanding application workloads, though, are highly variable or highly unpredictable, either because the activity that drives them is sporadic (such as end-of-month reporting cycles) or because user adoption can start small and then grow quickly.

Conventional hard-disk drive storage technology, for example, may not deliver sufficient performance for the kind of analytics that engineering and marketing departments need as they crunch larger and larger data sets. Solid-state drives (SSDs), on the other hand, eliminate the mechanical processes that slow down conventional drives — resulting in performance that is far superior.

However, IT decision-makers may be concerned about justifying the capital cost of an expensive SSD implementation.

The answer, suggests independent IT consultant Anil Desai, is to rent SSD capacity on an as-needed basis in the cloud.

“With the cloud, you can pay for just the capacity you need, just when you need it,” he says. “You can also spin up that capacity in a few minutes, instead of having to keep your business waiting for weeks or months while you bring it on line the way conventional data centers do today.”

Whose Cloud Is It, Anyway?

The term “cloud” is used broadly to refer a wide range of provisioning alternatives. Essentially, for a provisioning model to be considered cloud, it should possess three basic attributes:

1. Virtualization: Virtualization allows the computing, storage or network capacity of a physical device to be logically segmented and allocated to any designated workload, rather than fully dedicated to one application.

2. Automation: In a true cloud environment, virtualized resources are automatically — rather than manually — allocated to different workloads based on changing demand levels and other parameters.

3. Software-defined: To automatically allocate virtual resources, cloud environments must be controlled by software that can be programmed with appropriate rules and policies.

One way to categorize these alternatives is by the service being delivered. With Software as a Service (SaaS), an organization buys access to a specific application, such as Microsoft Office 365. The SaaS provider is responsible the entire underlying IT environment. With Platform as a Service (PaaS) offerings such as Microsoft Azure, an organization buys access to a complete IT environment — including databases, middleware and associated development tools — on which it can run applications it licenses or develops itself. And under an Infrastructure as a Service (IaaS) model, an organization buys access to foundational IT resources, such as servers, storage, networking and firewalls. IaaS vendors typically also offer value-added management services.

A variety of other cloud offerings are available as well, including databases, analytics, mobile device management and more. Even web hosting can be viewed as a cloud offering. What these all have in common is the offloading of technology ownership tasks to a vendor and payment on a subscription basis, rather than as a capital purchase. The cloud, however, can be more comprehensively categorized according to where and how application workloads are provisioned. This categorization typically includes:

• Public cloud: Under this model, technology resources reside in the service provider’s facility and are pooled across all users in what is referred to as a multitenant environment, which means that they are logically portioned to shield users’ data and application instances from each other.
• Private hosted cloud: This model is similar to a public cloud, except the service provider dedicates separate resources for each customer rather than using a multitenant approach. Some SaaS and PaaS providers offer single-tenant solutions. However, the private hosted model is most often associated with IaaS offerings.
• Private on-premises cloud: Under this model, a company creates a cloud environment within its own data center, dynamically allocating resources to various application workloads as necessary to fulfill constantly shifting demands for scale and performance.
• Hybrid cloud: Rather than provision every application in the same manner, many organizations adopt a hybrid cloud model under which they mix and match public and private alternatives as appropriate given considerations of cost, scale, performance, compliance, security and technical suitability.

“Most companies will move gradually to the cloud and will adopt different provisioning models for their different business needs,” says CDW’s David. “So just about everyone’s data center is eventually heading for a hybrid cloud.”

Learn more about CDW's solutions for the end of Windows 2003 support here.

(Samarets/ThinkStock)

The Data Center Payoff

Of course, data center modernization is about much more than just navigating the end of Windows Server 2003’s useful life or expediently hopping to another platform. It’s about having a long-term strategy for maximizing the ability of the enterprise to leverage technology as a means of sustainable competitive advantage.

And that, ultimately, is what makes the move to cloud so compelling. By migrating the right workloads to the right cloud provisioning model, IT decision-makers can gain multiple high-value benefits, including:

Reduced costs: The cloud dramatically reduces capital budget requirements by moving IT to a pay-as-you-go subscription model. It also reduces operating expenses by taking advantage of service providers’ economies of scale.

Faster responsiveness: Conventional data centers slow an IT department’s ability to add capacity and implement new applications. With clouds, IT can spin up many new capabilities in just minutes.

Scalable performance: Cloud data centers enable IT staff to effortlessly handle both predictable and unpredictable demand spikes that would otherwise adversely affect the organization, while eliminating the need to wastefully over-provision capacity when it’s not needed.

Business resiliency: Because cloud environments are automated and software-defined, they can quickly and easily be replicated to alternative facilities in the event of a disaster. This helps safeguard the continuity of critical operations under all kinds of potentially disruptive conditions.

More innovative IT: As IT departments automate low-value operational tasks or offload them to cloud providers, IT staff can focus more on strategic innovation that delivers greater advantages to the organization. The bottom line: IT decision-makers can take advantage of the end of Microsoft’s support for Windows Server 2003 to initiate a move to the cloud that will benefit the enterprise for many years to come.