Browse Prior Art Database

Adaptive Offset Setting for Post-Write Error Retry

IP.com Disclosure Number: IPCOM000118950D
Original Publication Date: 1997-Sep-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 4 page(s) / 81K

Publishing Venue

IBM

Related People

Hashimoto, M: AUTHOR [+2]

Abstract

Disclosed is a method of implementing a Post-Write Error Retry procedure without adding any complicated hardware or software resources. By changing the method of pacing the host data transfer, the post-write error retry can easily be achieved without degrading random write-cache performance without any micro code intervention.

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

Adaptive Offset Setting for Post-Write Error Retry

      Disclosed is a method of implementing a Post-Write Error Retry
procedure without adding any complicated hardware or software
resources.  By changing the method of pacing the host data transfer,
the post-write error retry can easily be achieved without degrading
random write-cache performance without any micro code intervention.

      The Post-Write Error Retry is a scheme to ensure that write
data operation is done without any error.  Among the error causes of
write operation, there are some that cannot be detected during the
operation.  The external mechanical shock is the most suitable
example of this type of error cause for a Hard Disk Drive (HDD) which
is employing  the embedded sector servo.

      With the embedded sector servo, the position of the writing
head is measured periodically from the servo patterns on the disk.
If an external mechanical shock causes a position of the head to
shift from  the target, the position error cannot be detected until
the next servo  pattern is read (Fig. 1).

      Usually, a small HDD uses a ring sector buffer and the Host can
fill new data soon after the old data is written onto the disk.  With
this type of pacing scheme, even if an error is detected at the next
coming servo pattern, no retry operation can be done because the data
on buffer is already written over by host.  To avoid this, there
should be some "offset" between the disk buffer pointer and host
buffer pointer.  With this offset, the data to be written is secured
at least  for this offset number of sect...