Browse Prior Art Database

Method for Recovering from Runaway Sector Reassignments on Hard Disk Drive

IP.com Disclosure Number: IPCOM000122596D
Original Publication Date: 1991-Dec-01
Included in the Prior Art Database: 2005-Apr-04
Document File: 2 page(s) / 106K

Publishing Venue

IBM

Related People

Forrer Jr, TR: AUTHOR

Abstract

This invention pertains to any Direct Access Storage Device (DASD) that communicates to the host system via an intelligent command level interface of such a type that, when requesting data, the host uses Logical Block Addressing (LBA) that contains no physical information as to where the data is located (i.e., cylinder, head or sector number). When LBA addressing is utilized by the host system, the device can perform defect management transparent to the host. Examples of these types of interfaces are the ANSI SCSI 1 or 2 interfaces or the IPI Level 3 Interface.

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

Method for Recovering from Runaway Sector Reassignments on Hard Disk
Drive

      This invention pertains to any Direct Access Storage
Device (DASD) that communicates to the host system via an intelligent
command level interface of such a type that, when requesting data,
the host uses Logical Block Addressing (LBA) that contains no
physical information as to where the data is located (i.e., cylinder,
head or sector number). When LBA addressing is utilized by the host
system, the device can perform defect management transparent to the
host.  Examples of these types of interfaces are the ANSI SCSI 1 or 2
interfaces or the IPI Level 3 Interface.

      The SCSI interface shall be used to describe the application of
the invention.

      During the use of DASD, disk surface defects are occasionally
encountered during the course of writing and reading data to or from
the media.  Either one of two actions are taken to hide the defect
from the host.  DASD that support the automatic defect reassignment
capability will automatically reassign the defective location (LBA).
For DASD that does not support automatic reassignment, the host
issues the Reassign Block Command and then the DASD reassigns the
LBA.  In both cases, the DASD hides the bad LBA from the system and
puts the physical address of the defect in a grown defect list called
the Glist.

      During the use of the DASD, either external system conditions,
such as excessive shock, vibration, both internal (to the system)
and/or external (surrounding environment), electronic or magnetic
field interference caused by other system component failures, such as
displays or power supplies, can cause the DASD to not be able to read
data.  Also internal DASD component failures can cause data loss.
All of these conditions result in LBA reassignment. Many system's RAS
Strategy, including the RISC System/6000*, offer the customer the
option of repairing the DASD by replacing the drive electronics card.
This repair process is done to help the customer recover their data.
However, in doing this, bad data previously written to the media is
still there.  These bad LBAs will be reassigned as they are
encountered.  If enough LBAs were reassigned, this would ultimately
result in all of the spare LBAs (each DASD has a finite number of
spare LBAs designated for reassignment purposes) being used.  When
this happens the entire drive assembly must be replaced or
reformatted using the options for the Format Command that ignore all
of the defects found during customer use (Glist defects).  That is,
only the defects found during the DASD manufacturing process
(contained on the DASD in what is...