KIRK HANSON
Architects at Pajaro Way, Pearl, ID

6553509, Apr 22, 2003

Jul 28, 1999

09/363917

Kirk A. Hanson - Eagle ID
Douglas L. Voigt - Boise ID
Don L. Burkes - Meridian ID

Hewlett Packard Development Company, L.P. - Houston TX

H02H 305

714 5, 714 20

A storage system maintains a log image consisting of log records kept in a page log and log records kept in a distributed log kept in staging buffers of multiple storage media. The distributed log contains incremental log records that are occasionally forced to the storage media prior to filling an entire page of log records. The storage system has a log recovery manager that recovers the log image following a failure. The log recovery manager first reads the log records from the page log to reproduce a majority of the log image. The log recovery manager then scans the distributed log to locate any more recent log records that may exist. Once a more recent log record is found, the log recovery manager adds it to the recaptured log image and proceeds to find even more recent log records. To speed the recovery, the log recovery manager intelligently begins the search for the next incremental log record at a location in the storage system that the next log record is likely to reside. The log recovery manager begins looking for the next log record using three criteria: (1) it looks on the same storage medium that contains the previous log record just found; (2) it looks in the other staging buffer on the storage medium; and (3) it begins at an offset equal to the length of the previous log record just found.

6055604, Apr 25, 2000

Aug 26, 1997

8/920120

Douglas L. Voigt - Boise ID
Don L. Burkes - Meridian ID
Kirk A. Hanson - Eagle ID

Hewlett-Packard Company - Palo Alto CA

G06F 1208

711117

A transaction log stored in a first memory is posted to one of two separate log areas, selectively, on a storage system having a plurality of storage media such as disk drives. Posting occurs to a "disk log" area when a page-full status of the first memory transaction log is detected. Alternatively, when a posting request occurs before a page-full status of the transaction log has been detected, then posting occurs immediately to a least busy disk of a "staging log" area. Posting occurs to the least busy disk of the "staging log" area by circumventing normal storage system data management and redundancy techniques such that data in the staging log area is not kept redundantly on the storage media. Rather redundancy is maintained by the fact that the transaction log remains with the first memory in addition to being copied to the staging log area. The staging log area includes reserved space on each of the plurality of storage media and is divided into logically separate portions on each of the storage media.