Browse Prior Art Database

Hardware and Volume Statistical Analysis and Reporting System for Detecting and Isolating Media and Hardware Errors on the Magstar Family of Tape Drives

IP.com Disclosure Number: IPCOM000013086D
Original Publication Date: 1999-Dec-01
Included in the Prior Art Database: 2003-Jun-12
Document File: 3 page(s) / 52K

Publishing Venue

IBM

Related People

M. E. Hill: AUTHOR [+6]

Abstract

Hardware and Volume Statistical Analysis and Reporting System for Detecting and Isolating Media and Hardware Errors on the Magstar Family of Tape Drives In tape libraries and data processing centers, even a few bad tapes can cause many problems. One problem is that these tapes exhibit temporary and/or permanent errors during reading and/or writing. These errors reduce system throughput and jeopardize the integrity of the data on the tape. Another problem is that bad tapes sometimes deposit debris on the heads of the tape drives in which the tapes are mounted. Subsequently, this debris is then deposited on other tapes, which then deposit the debris on to heads of other tape drives, etc. When enough debris accumulates on the head of a tape drive, the drive exhibits temporary and/or permanent errors during reading and/or writing of tapes. As stated above, these errors reduce system throughput and jeopardize the integrity of the data on the tape. Furthermore, customer service engineers must then clean the heads of tape drives in which the bad tapes were mounted. This is a time consuming, manual process which costs the customer money, both for a service call and system down time.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 21% of the total text.

Page 1 of 3

Hardware and Volume Statistical Analysis and Reporting System for Detecting

and Isolating Media and Hardware Errors on the Magstar Family of Tape Drives

In tape libraries and data processing centers, even a few bad tapes can cause many problems. One problem is that these tapes exhibit temporary and/or permanent errors during reading and/or writing. These errors reduce system throughput and jeopardize the integrity of the data on the tape.

Another problem is that bad tapes sometimes deposit debris on the heads of the tape drives in which the tapes are mounted. Subsequently, this debris is then deposited on other tapes, which then deposit the debris on to heads of other tape drives, etc. When enough debris accumulates on the head of a tape drive, the drive exhibits temporary and/or permanent errors during reading and/or writing of tapes. As stated above, these errors reduce system throughput and jeopardize the integrity of the data on the tape. Furthermore, customer service engineers must then clean the heads of tape drives in which the bad tapes were mounted. This is a time consuming, manual process which costs the customer money, both for a service call and system down time.

To minimize their effects, bad tapes must be isolated and discarded as quickly as possible. Currently, isolating these bad tapes is a challenging and time consuming problem for customer service engineers.

In addition, some defects in hardware may degrade the system so that processing tape becomes harder. Small defects that are easily corrected, or corrected as the error occurs, may indicate a hardware failure. These types of errors can also jeopardize throughput and reduce margin in tape drives, making more serious errors more likely to happen. This loss of margin or small loss of performance, if detected, can consume many service hours in isolating the problem manually. To minimize this loss of margin and/or throughput, these errors must be detected, isolated, and reported as quickly as possible.

A statistical methodology can be used to isolate between hardware and media errors on tape drives, and to predict future errors. This includes analyzing drive history for hardware errors, media history for media errors, and combining the two to isolate failures. In IBM 3590 and 3570 tape drives, this statistical method is called SARS--Statistical Analysis and Reporting System.

This publication describes the SARS data used in IBM 3590 and 3570 tape drives. It does not detail the algorithms or thresholds used in the algorithm.

Previous products have used a SARS algorithm. These were implemented using information at the host (or at a control unit, which appears as a host to a drive). In the IBM 3590 and 3570 tape drives, SARS is implemented in the drive. Both methods store information about mounts of a tape cartridge in a drive. The host method stores information at the host, while the proposed drive method stores information both in the drive and the tape cartridge. Ev...