
Implementing a Tiered Storage Strategy
By Alice LaPlante
The cost of storage continues to decline even as the capacity of individual drives increases. But rather than move all data to the fastest -- and most expensive -- devices, most enterprises are implementing a tiered storage strategy, in which data is delegated to devices according to its importance.
"Ten years ago, people simply installed systems that allowed them to store all information electronically and to get to it quickly," says Joe Martins, a partner with the Data Mobility Group LLC, in Nashua, N.H. "They didn't start to consider strategies that took into account cost as well as performance issues until 2000," when resolving Y2K issues brought it to their attention.
Dividing the Data
The Enterprise Strategy Group estimates that about 85 percent of the data sitting on high-end storage systems is non- or post-transactional, and that half of this replicates non-transactional data. "This is a very inefficient use of high-end real estate at best -- and a costly business adventure at worst," says Heidi Biggar, an analyst with the Enterprise Strategy Group, based in Milford, Mass. After all, primary disk storage is expensive to acquire, manage and maintain, as well as draining power and cooling resources.
Best practices involve first classifying data into categories. Information deemed mission-critical and data that is frequently accessed and/or changed should be (article continues)
Next Page >>