Autonomic retention classes

• US 7,552,131 B2
• Filed: 03/05/2007
• Issued: 06/23/2009
• Est. Priority Date: 03/05/2007
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation, National Security Agency
• Original Assignee: International Business Machines Corporation
• Inventors: Uttamchandani, Sandeep M., McCloskey, Joseph P., Chron, Edward G., Pease, David A., Richards, Elizabeth S.
• Primary Examiner(s): Pardo; Thuy N
• Application Number:
  US11/681,955
• Publication Number:
  US 20080222075A1
• Time in Patent Office: 841 Days
• Field of Search: 707/7, 707/101, 707/102, 707/103.Z, 707/104.1, 707/204, 715/206, 711/202, 711/216, 382/159
• US Class Current: 1/1
• CPC Class Codes: G06F 3/0652 Erasing, e.g. deleting, data cleaning, moving of data to a wastebasket
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: Y10S 707/99937 Sorting
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface
• CPC Class Codes: Y10S 707/99944 Object-oriented database structure
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

The embodiments of the invention provide methods, computer program products, etc. for autonomic retention classes when retaining data within storage devices. More specifically, a method of determining whether to retain data within at least one storage device begins by storing data items in at least one storage device. Furthermore, the method maintains access statistics for each of the data items, an age of each of the data items, and an administrator-defined importance value of each of the data items. Following this, a retention value is calculated for each of the data items based on the access statistics for each of the data items, the age of each of the data items, and the administrator-defined importance value of each of the data items.

Claims

1. A method of determining whether to retain data within at least one storage device, said method comprising:

storing data items in at least one storage device;

maintaining access statistics for each of said data items;

maintaining an age of each of said data items;

maintaining an administrator-defined importance value of each of said data items; and

calculating a retention value for each of said data items based on said access statistics for each of said data items, said age of each of said data items, and said administrator-defined importance value of each of said data items;

wherein said calculating of said retention value comprises creating clusters of data points, wherein each of said data points comprise a function of data item age and data item number of accesses; and

wherein said calculating of said retention value further comprises mapping said clusters into linear functions;

wherein said mapping of said clusters into said linear functions comprises creating said linear functions such that said linear functions comprise slopes that are inversely proportional to densities of said clusters;

wherein said mapping of said clusters into said linear functions comprises creating said linear functions such that coordinate ranges of said linear functions correspond to diameters of said clusters;

wherein said calculating of said retention value further comprises combining said linear functions with said administrator-defined importance value of each of said data items in an adjustable weighted function; and

wherein said removing data items that have a retention value below a predetermined value from said storage device.