US 7,464,250 B2
Method to reduce disk access time during predictable loading sequences
Richard Alan Dayan, Raleigh, N.C. (US); and James Franklin Macon, Jr., Raleigh, N.C. (US)
Assigned to International Business Machines Corporation, Armonk, N.Y. (US)
Filed on Mar. 11, 2004, as Appl. No. 10/798,091.
Prior Publication US 2005/0204095 A1, Sep. 15, 2005
Int. Cl. G06F 9/06 (2006.01)
U.S. Cl. 711—217  [711/112; 711/173; 711/213; 711/218] 1 Claim
OG exemplary drawing
 
1. A method of loading data from disk in a data processing system, comprising:
comparing a current sequence of disk requests to data indicative of a previous sequence of disk requests, wherein the sequence of disk requests includes the sequence of disk requests following a power-on event;
responsive to detecting a match between the current sequence and the previous sequence, storing a copy of data blocks accessed during the current sequence in a contiguous portion of the disk, wherein storing a copy of data blocks accessed during the I/O sequence comprises storing the data blocks sequentially in the order that the data blocks were accessed chronologically; and
responsive to a subsequent request for data in the disk sequence, mapping the request to the sequential portion of the disk and servicing the request from data in the sequential portion;
wherein the contiguous portion of the disk to which the data is copied is on a different partition of the disk than a disk partition on which the original data is stored;
recording a sequence of disk accesses, wherein recording the sequence includes recording the disk address of each block accessed and the length of each block;
responsive to retrieving data from the contiguous portion, prefetching additional data from the contiguous portion of the disk and caching the additional data in a buffer; and
responsive to an I/O request, determining whether the data requested resides in the buffer and, if so, retrieving the data from the buffering without accessing the hard disk.