Browse Prior Art Database

Predictive Failure Analysis of Reassign Rate

IP.com Disclosure Number: IPCOM000108788D
Original Publication Date: 1992-Jun-01
Included in the Prior Art Database: 2005-Mar-22
Document File: 2 page(s) / 57K

Publishing Venue

IBM

Related People

Hicken, MS: AUTHOR [+5]

Abstract

A method for alerting customers to media problems on direct-access storage devices (DASDs) is disclosed. The method is called reassign rate predictive failure analysis (PFA).

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 72% of the total text.

Predictive Failure Analysis of Reassign Rate

       A method for alerting customers to media problems on
direct-access storage devices (DASDs) is disclosed.  The method is
called reassign rate predictive failure analysis (PFA).

      By tracking media defects which have caused significant amounts
of time to be spent in data recovery, an increase in the probability
of data loss can be detected.

      By applying limits to the rate at which significant media
defects develop, the point at which this failure mechanism is
established can be reported.

      This method of predictive failure analysis is used to alert the
customer to media problems which may affect the device's ability to
recover data in the future.

      The Media Error Log contains several defect locations and the
associated reassign scores.  The defect location indicates where the
defect is located on the device relative to some reference point.
The reassign score is related to the severity of the error.  Each
time the defect location is read and data recovery is required, the
media error log accumulates the reassign score.

      The reassign threshold is established on the reassign score
such that, if exceeded, it would indicate that the defect has become
severe enough to warrant reassignment. Reassignment requires that the
data contained in the media location be moved to a different location
on the media so that the defect no longer poses the risk of customer
data loss.  The d...