Tackling Storage Overload Via Information Lifecycle Management Part 1 Data Classification

Tackling Storage Overload Via Information Lifecycle Management, Part 1: Data Classification

By Joe Purcell
Expert Author
Article Date: 2011-07-11

"We are drowning in a sea of information. No one knows the value, how long it should be kept, or what the risk of keeping or getting rid of it is" (J.P. Morgan). A white paper by Recommind on getting data under control mentions that "the amount of enterprise data doubles every 18 months," according to an IDC report. Be it an individual or massive organization, hours can be wasted on trying to find information and nightmares can plague trying to manage it if a solid Information Lifecycle Management (ILM) policy isn't made. An ILM plan begins with defining the data's properties, category, and type.

There are three main properties of any piece of data, as one article states: Confidentiality, Integrity, and Availability (CIA). The article explains that Confidentiality is the need to place strict access control, Integrity is the need for accuracy of the data, and Availability is the need for the data to be accessible at all times. Every piece of data can be then classified as Level 1, 2, or 3 using the following table:

	Level 1	Level 2	Level 3
Confidentiality	Required	Recommended	Optional
Integrity	Required	Recommended	Optional
Availability	Required	Recommended	Optional

Throughout the lifecycle of a piece of data it will shift in each CIA category, and the management of that data will shift as well. If a piece of data is highly confidential, but is older than 10 years and hasn't been accessed in that time then it can be moved from hard disk storage to tape storage for archiving. The CIA properties determine how the data should be stored, but to determine what properties the data has we'll need to know what category the data is in.

Categorizing data can be very difficult. The best way to begin is to write down all the data in question in a hierarchical manner as it currently is, even if it is unorganized. Then begin to sketch what network or category the data is in. Using a network-based categorization can be simply External and Internal, or by department such as Treasurer or Operations. Otherwise, data can be classified by made up categories like finance or media.

Finally, label what data types are in each category. For example, finance will have many small text-based files, such as PDF, Word, and Excel. Media will have fewer large files, such as MP3 or VOB. Knowing what file types are in the category will shape how that data is managed. For example, one would want to put large media files in an archived place more quickly because of how much space they take up, whereas finance data not as much because it is smaller.

Once all your data is labeled and properties attached, then a scheme for storing the data can be made. Reorganizing the list of the current data structure may take a few iterations. The first round may involve focusing on just media to free up more storage space. The next round may involve moving sensitive data onto an encrypted volume. Companies often fail to look at their ILM policy and in doing so waste a lot of needed storage space and, in the worst case scenario, face fines for not having needed information.

About the Author:
Joe Purcell is a technology virtuoso, cyberspace frontiersman, and connoisseur of Linux, Mac, and Windows alike.