The logical place to start is with the data and to classify the data according to type. However, says Richard Martin of Hopkinton, MA-based EMC, the best place to start is with the applications.
Martin, global program manager for EMC's Information Consulting group oversees EMC's ILM consulting practice. In the company's experience, he says, applications should lead the data.
"We use applications as the organizing principle," Martin says. "It's practical and gives fast benefits as opposed to a more granular review of data types."
While classification by data type is a more theoretically satisfying solution, it is also harder to implement, in part because of ambiguities inherent in data classification. Classification by application is a quick-and-dirty approach that lets ILM start paying dividends almost immediately. The reason, Martin says, is that classification is not just a guide to handling data throughout its life, it also helps determine what kind of storage should be used for the data in the first place.
One of the factors that makes ILM attractive is the availability of different classes of storage, well beyond the traditional disk and tape dichotomy. Today there are several different classes of each kind of storage that trade price for speed or other characteristics. This allows enterprises to save considerable money by building tiered storage architectures that use several different kinds of disk storage to more cost-effectively handle their data.
However an effective tiered storage architecture depends on assigning data storage to the appropriate tier. "We think that application based classification for tiering is very practical," Martin says. "Beyond that, we think that data policy for targeted applications where we have to outline the overall life cycle is the next step."
For more information:Tip: Here comes ILM
Tip: The best practices of storage management
Tip: ILM standards war heats up
About the author: Rick Cook has been writing about mass storage since the days when the term meant an 80K floppy disk. The computers he learned on used ferrite cores and magnetic drums. For the last twenty years he has been a freelance writer specializing in storage and other computer issues.
This was first published in April 2004