Browse Prior Art Database

Method to Manage Sector Defects

IP.com Disclosure Number: IPCOM000062393D
Original Publication Date: 1986-Nov-01
Included in the Prior Art Database: 2005-Mar-09
Document File: 2 page(s) / 14K

Publishing Venue

IBM

Related People

Tomasetti, GA: AUTHOR [+2]

Abstract

A method is described to manage sector defects that occur on disk media. The method uses a sector re-mapping method to handle defective sectors. Sectors that are found to be defective either at manufacturing time or during use are recorded in a bad block map that is contained on cylinder zero of each drive. This map contains the location of all defective sectors on the drive as well as addresses of alternate sectors. This map is initialized at drive manufacture with a list of manufacturer-found defects. When the drive is installed, an alternate pool of blocks is allocated based on drive size and all known defects are then assigned an alternate sector. Any other defects that are found later will be assigned alternates from this pool as well.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Method to Manage Sector Defects

A method is described to manage sector defects that occur on disk media. The method uses a sector re-mapping method to handle defective sectors. Sectors that are found to be defective either at manufacturing time or during use are recorded in a bad block map that is contained on cylinder zero of each drive. This map contains the location of all defective sectors on the drive as well as addresses of alternate sectors. This map is initialized at drive manufacture with a list of manufacturer-found defects. When the drive is installed, an alternate pool of blocks is allocated based on drive size and all known defects are then assigned an alternate sector. Any other defects that are found later will be assigned alternates from this pool as well. To minimize the path length associated with requests to the media, the sector defect map is only read when a minidisk is opened. The map is then parsed to yield a list of defects for the minidisk. This allows a smaller list to be looked at when the minidisk is accessed and also saves memory because the whole sector defect map is not contained in memory. Accesses to the media that do not use minidisks do not have any bad block management support since access by minidisk is the usual method used for data access. All accesses to a minidisk must be compared with the in- memory list of defects. This is done so that if a defective sector is accessed, the arm can be moved to the alternate sector to access the data. This lookup could have contributed to the path length of a request. Therefore, the lookup is done in parallel to activity being done on the drive for other requests. If a sector defect is contained in the request, an indication is added to the request to inform the disk device driver that special processing must occur for this request. This prevents any path length additions from occurring during device driver processing of the request in the usual case of no defects. Additional interfaces are provided to the disk device driver by the minidisk manager to facilitate defective sector processing. The module $CHKBLK is called if the bad block indication is active in the re...