August 12, 2002

The Next IT Imperative

As storage investments grow, sound information infrastructure will rely heavily on a solid storage strategy

by Richard Scannell

Historically, most organizations purchase storage assets as a component of vertical projects. From the Web rush and Y2K threat of the late 1990s, to the supply chain and CRM initiatives of more recent years, storage has been viewed as just another element of the overall architecture.

Although this approach was well and good in the past, new strategic business applications such as state-of-the-art enterprise applications and business intelligence solutions create a greater demand for storage, leading to enterprise storage needs that often double in size annually. This vertical approach leads to significant installed bases of storage, and without a rationalized approach to managing and growing it, these applications will not scale effectively and certainly not cost efficiently.

Within the average enterprise, spending on storage and its associated costs will soon outstrip all other IT costs. According to the Enterprise Storage Group, storage currently accounts for 70 percent of the typical IT capital budget. Any activity that consumes such a large percentage of IT spending must be approached strategically, prompting the enterprise to shift gears and approach storage as an initiative in its own right by setting goals that align business objectives with infrastructure solutions and fiscal expenditures. By raising the focus of storage to a level worthy of any large portion of the IT budget, organizations will begin to realize tremendous efficiency gains and dramatically improved operations.

As with any large initiative, the first task is to develop a strategy. Generally, the consensus is that the only effective way to scale to the levels of data that will be present in the future is through centralized storage. However, centralized storage doesn't suggest a specific technology solution — such as storage area networks, network-attached storage, or direct-attached storage — as much as it points to the need for an organizational approach.

With your strategy in place, your organization can develop a technology framework that supports user, financial, and IT needs. The components of such a storage strategy include the following:

• Inventory
• Data classification and occupancy
• Cost modeling
• Developing organizational competence
• Selecting as well as implementing the appropriate technologies
• Measuring and reporting.

Inventory

Your IT organization should approach the rationalization of storage the same way it rationalized the disparate applications associated with CRM or supply chain initiatives. First, you need to thoroughly understand what systems are already in place and what purposes they serve: Whose data is on each device, why is it there, and how does it support the business? This assessment is valuable not only in the front end of the strategy but also at the back end where data migration occurs. This inventory should cover not only hardware and software but also employee skill sets and existing internal processes and policies. With an inventory of current assets available, the path toward data classification becomes clearer.

Data Classification and Occupancy

In its simplest term, data classification divides data into three types: service-affecting data, internal-affecting data, and retained data. Service-affecting data is mission critical and directly affects the ability of your business to deliver to its customers should it become unavailable or irrecoverable. Internal-affecting data, if unavailable, harms the internal operations of the business but doesn't affect customers. Retained data must be stored but will rarely be accessed — for example, regulatory requirements in the healthcare industry dictate that patient medical records must be retained for a specified number of years. However, it's important to note that for any given business, fewer or more classifications may be necessary.

Occupancy is the measure of how much data falls into each classification. Occupancy shouldn't be confused with utilization, which relates to storage resource management and virtualization and their ability to improve asset value from 30 percent utilized to 60 percent utilized via improved intelligence and device allocation. In contrast, determining occupancy levels will lead to the conclusion that an organization utilizing 90 percent service-affecting data and 10 percent retained data will have a very different cost of ownership scenario than an organization with just 20 percent service-affecting data and 80 percent internal-affecting data.

Cost Modeling

You can use cost modeling to help determine a series of infrastructure standards for storage; typically, one standard for each data classification. Because commercially available products develop so quickly, enterprisewide standardization is difficult. For the purpose of cost modeling, defining architecture attributes is a much more rational approach.

• Privacy
• Your California Privacy Rights
• Copyright © 2008 United Business Media Limited