A server-based application computing solution can dramatically reduce application maintenance and desktop costs. Enterprises that consider this can also make its network less complex. by Soutima S. Das

Ever since computing moved onto the desktop, giving access to all kinds of applications to all categories of users in an organisation, an unrelenting complexity has dogged technology personnel, from the top IT honcho down to the system administrator. The complexity gives IT staff management nightmares, and top management the bugbear of ballooning costs, as they grapple with the nagging issue of expensive upgrades.

The complexity also involves a large number of applications written for stubbornly different operating systems running on a wide array of hardware platforms. In short, an expensive complexity that simply refuses to go away.

The issue of the rising cost of managing an intriguing Web of technical infrastructure has surfaced frequently in various industry surveys. Product

companies, both hardware and software, continue to roll out faster machines and software often with incremental value propositions, yet necessitating frequent upgrades in user organisations for competitive reasons.

Combine this with organisational churn and growth, and you see your goals of managing your computing infrastructure in today's tight budgets slipping out of your grasp.

Computing Models
While the issues of key hardware are mostly settled when an organization decides on a computing architecture across its operations, those associated with software upgrades, and application deployment & management are the most daunting. Today, more than ever, organizations seek to minimize the costs associated with owning, deploying, and managing applications. Organizations today typically use three primary computing models to deploy applications. These models are the unmanaged desktop, managed desktop, and the server-based application computing environment.

The unmanaged desktop computing model defines an environment in which users have their own desktop systems. System administrators manage each system individually, including the installation and upgrade of operating systems and software.

The managed desktop model allows administrators to install and manage applications on desktops remotely and from a centralized location.

The application computing environment is focused on a server-based computing model. In this environment, all application processing takes place on centralized servers. Users can access these applications, either locally or remotely, via virtually any connection using a standard desktop device, a thin client appliance, a wireless device, and a Web browser.

Cost Issues
Each of these models presents unique cost and support implications. Typically, the unmanaged desktop is the costliest, and server-based computing offers the lowest desktop management costs. To complicate matters, most organizations currently employ more than one of these computing models. This not only increases overall cost, but also creates technical delivery challenges for users who need access to applications that run under different models. Since these organizations also need to support multiple platforms like Windows, Unix, and Mac, it complicates matters further and dramatically increases costs under some of these models.

According to a study of these computing models, conducted by Doculabs, it has been found that the major cost components associated with deployment of any of the computing models are loss of productivity, desktop infrastructure, application maintenance, and training. Loss of productivity is usually due to server, desktop, and network downtime. Desktop infrastructure includes elements like hardware, operating system licenses, and support. And application maintenance includes desktop and client/server applications.

A server-based application computing solution dramatically reduces the costs associated with these key components, particularly application maintenance and desktop costs. In fact, it has been shown in the study that typical large organizations can save over 50 percent in hard costs in the first year following deployment.

Most organisations do not directly consider as part of their IT cost structures the costs associated with loss of productivity and training, which are often categorized as 'soft' costs. However, for long-term benefits, organizations must consider soft costs when determining how they will deploy applications in future.

Soft costs refer to costs associated with non-tangible items. These include the costs associated with loss of productivity due to downtime or training. And hard costs typically refer to the cost of items like network hardware, telecommunications, and software licenses. These are budgeted items that are clearly visible on a company's balance sheet.

Although cost is a measurable and well-understood concept, it can be difficult to calculate, especially in light of the number of variables an application infrastructure presents. An accurate assessment of cost is also the key to calculate other metrics like return on investment and profitability. Many variables influence the cost of deploying and managing applications. Some of these elements are hard numbers that are easy to ascertain, like the number of users and applications and the price of thin-client workstations. Other numbers are more difficult to determine, like the effects of downtime on different user populations.

Variables like the number of remote locations that need to be managed, vary greatly from organization to organization. Though it is difficult to take every element into consideration when calculating the cost, it is possible to use a smaller set of variables known to significantly influence cost. Understanding how these variables affect cost allows organizations to intelligently and effectively deploy applications.

Benefits of Server-Based Model
Once organisations have understood the impact of a carefully done cost analysis before selecting a computing model or migrating to another one, they can visualise the other benefits that accrue from a server-based computing environment. These benefits typically include:

• Improved centralized management of IT infrastructure.
• Unification of disparate and fragmented network infrastructure.
• Demonstrated cost benefits for organizations of all sizes that increase with the size of the organization.
• Seamless desktop integration of the user's local and remote resources.

Using a server-based application computing model, most organizations should see large returns in the first year of deployment. They will also find that support and desktop-related costs drop off quickly. Other major savings will result from reduced application downtime and diminished need for user support and training.

Generally, customers of server-based computing cite ease of deployment, reduced overall cost, and centralized management and control among the top benefits they have derived.

Organizations can realise significant benefits by moving to a server-based solution, particularly those relying on client/server applications to conduct business. In this day of tighter spending and more focused technology investment, server-based computing keeps costs low and manageability high. Customers will be able to mange their bottomlines more efficiently.I certainly see a shift to server-based computing among enterprises who are serious about its budgets. In the maze of expensive complexity, that should come as an easy way out.

Soutima S. Das is Managing Director, Citrix Software India