Browse Prior Art Database

Storage Data Revival System without Extra H/W

IP.com Disclosure Number: IPCOM000123589D
Original Publication Date: 1999-Jan-01
Included in the Prior Art Database: 2005-Apr-05
Document File: 1 page(s) / 46K

Publishing Venue

IBM

Related People

Sugawara, T: AUTHOR

Abstract

One of the most important problem on a storage device is data loss by a hard error on reading a sector. Data corruption can be easily occurred by any factors like a particle or a defect on media. Most storage device like Hard Disk Drive (HDD) has ECC bytes on every sectors, and is able to correct a certain length of data corruption on its medium by using them. However if the data corruption is too big to recover data, the data of the sector is lost forever.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 61% of the total text.

Storage Data Revival System without Extra H/W

   One of the most important problem on a storage device is
data loss by a hard error on reading a sector.  Data corruption can
be easily occurred by any factors like a particle or a defect on
media.  Most storage device like Hard Disk Drive (HDD) has ECC bytes
on every sectors, and is able to correct a certain length of data
corruption on its medium by using them.  However if the data
corruption is too big to recover data, the data of the sector is lost
forever.

   Our method is able to revive data of a lost sector on any
kind of Writable storage device, such as a HDD, CD-R, etc..

   FAT (File Allocation Table) file system is adopted on DOS,
MS-Windows, etc. for storage.  Data is stored and read at a Cluster
unit, which is constructed with several sectors like 8, 16, 32
sectors or etc..  The method consists of following:
  1.  A cluster of FAT is constructed with a certain number
      of data sectors and one parity sector which doesn't
      contain normal data.
  2.  When a cluster of data are stored on the device,
      A parity sector is produced   from the data sectors
      in the cluster.  The data sectors and the parity sector
      are stored on the device as one cluster.  An example of
      the way to produce a parity sector is to calculate Nth
      byte of the parity sector by executing XOR among each
      Nth bytes of the data sectors.
  3.  When to read...