Richard J Defouw

US Patent:

6742084, May 25, 2004

Filed:

May 4, 2000

Appl. No.:

09/564261

Inventors:

Richard J. Defouw - Boulder CO
Alan Sutton - Boulder CO
Ronald W. Korngiebel - Westminister CO

Assignee:

Storage Technology Corporation - Louisville CO

International Classification:

G06F 1200

US Classification:

711133, 711134, 711135, 711136, 711171

Abstract:

A caching method for selecting variable size data blocks for replacement or removal from a cache includes determining the size and the unreferenced time interval of each block in the cache. The size of a block is the amount of cache space taken up by the block. The unreferenced time interval of a block is the time that has elapsed since the block was last accessed, and may be determined using a least recently used (LRU) algorithm. The recall probability of each block in the cache is then determined. The recall probability of a block is a function of its unreferenced time interval and possibly size and other auxiliary parameters. The caching method then determines a quality factor (q) for each block. The (q) of a block is a function of its recall probability and size. The caching method concludes with removing from the cache the block with the lowest (q).

US Patent:

7139863, Nov 21, 2006

Filed:

Sep 26, 2003

Appl. No.:

10/672786

Inventors:

Richard John Defouw - Boulder CO, US
Thai Nguyen - Thornton CO, US

Assignee:

Storage Technology Corporation - Louisville CO

International Classification:

G06F 12/00

US Classification:

711103, 711165, 36518533

Abstract:

A method, system and apparatus for improving the useful life of non-volatile memory devices such as flash memory. The present wear-leveling technique advantageously improves the overall useful life of a flash memory device by strategically moving inactive data (data that has been infrequently modified in the recent past) to the memory blocks that have experienced the most wear since the device began operation and by strategically moving active data to the memory blocks that have experienced the least wear. In order to efficiently process and track data activity and block wear, vectors of block-descriptor pointers are maintained. One vector is sorted in decreasing order of overall block erase/write activity (block-wear indicator), whereas the other vector is sorted in increasing order of the number of times a block has been erased since the last wear-leveling event occurred (activity indicator for the data stored in the block). The activity levels of the data and the wear levels of the blocks are then easily compared and otherwise processed using pointers into these vectors to allow for more efficient processing than previous techniques used for wear leveling.

US Patent:

7437511, Oct 14, 2008

Filed:

Jan 18, 2006

Appl. No.:

11/334127

Inventors:

Thai Nguyen - Thornton CO, US
Michael L. Leonhardt - Longmont CO, US
Richard John Defouw - Boulder CO, US

Assignee:

Storage Technology Corporation - Louisvile CO

International Classification:

G06F 12/08

US Classification:

711120, 711168, 710 32

Abstract:

For use in a storage area network (SAN), a virtualization layer including at least one virtual engine having a respective local cache and a secondary cache layer, wherein the secondary cache layer includes the local caches coupled together, the local caches individually including a first cache layer, and at least one of a data transfer command and data corresponding to the transfer command are multicast to the secondary cache layer through an interconnection bus, the interconnection bus coupling the at least one virtual engine and at least one physical storage device.

US Patent:

55008038, Mar 19, 1996

Filed:

Nov 23, 1993

Appl. No.:

8/156493

Inventors:

Frederick Munro - Broomfield CO
Aaron Dailey - Boulder CO
Richard Defouw - Boulder CO
David Trachy - Louisville CO

International Classification:

G06F 1700
G11B 1568
B65G 100

US Classification:

364478

Abstract:

An automated cartridge system optimizes the time it takes to execute a series of cartridge requests. After cartridge requests are received, the library controller calculates the approximate time it will take to execute each possible sequence of pending cartridge requests. The library controller then executes the first request in the sequence of pending requests which will take the shortest elapsed time. This series of calculations and executions continues until all requests are completed. It is another object of the present invention to carry out the motion optimization in such a manner that assures that a particular cartridge request is not put off indefinitely. Each time a hand or arm is allocated to a request, all other requests waiting for these mechanisms have their calculated execution times reduced to increase the likelihood of execution.