Method and system for evaluating intrusive repair for plurality of devices

• US 6,009,246 A
• Filed: 01/13/1997
• Issued: 12/28/1999
• Est. Priority Date: 01/13/1997
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Kanthanathan, Manoranjan, Grzinich, John Carl, Chandra, Arun
• Primary Examiner(s): Beausoliel, Jr., Robert W.
• Assistant Examiner(s): HUA, LY
• Application Number:
  US08/783,595
• Time in Patent Office: 1,079 Days
• Field of Search: 395/183.13, 395/183.02, 395/183.21, 395/183.01, 395/182.04, 395/182.09, 371/26, 371/3, 364/551.01, 364/580
• US Class Current: 714/26
• CPC Class Codes: G01R 31/31835 Analysis of test coverage or failure detectability
• CPC Class Codes: G06F 11/006 Identification G06F11/2289 takes precedence

Abstract

The present invention discloses a system and method to evaluate intrusive repair. The method and system is based on analyzing data from the field and drawing causal diagrams (cause-consequence diagrams). In an aspect of the present invention, a method and system for evaluating intrusive repair for a class of devices comprises providing a threshold time window for analyzing failure data. The system and method further includes utilizing a causal diagram based on the analysis of the failure data for the class of devices to indicate the probability of intrusive repair. In the present invention, two measures are defined to evaluate intrusive repair--intrusivability and one-fixability. Intrusivability is defined as the probability that a repair action will cause a new fault. One-fixability is defined as the probability that the diagnostic process will fix the fault on the first repair attempt. Both of these measures are determined by using the causal diagram in accordance with the present invention.

Claims

1. A method for evaluating intrusive repair for a class of devices comprising the steps of:

providing a threshold time window for analyzing failure data; and

utilizing a causal diagram based on the analysis of the failure data for the class of devices to indicate the probability of intrusive repair determined by a measurement of intrusivability and one-fixability.

2. The method of claim 1 in which the failure data is categorized as classes of faults which are identified utilizing error log analysis techniques.

3. The method of claim 2 in which classes of devices comprise computer processing systems.

4. The method of claim 3 in which classes of faults comprise CPU faults, memory faults, power supply faults, software faults and I/O faults.

5. The method of claim 1 in which the causal diagram comprises:

first path wherein a device after a first fault is repaired requires no further repair action;

a second path wherein the device after a first fault is repaired encounters a first type of fault, the first type of fault being different than the first fault (second fault); and

a third path wherein the device after the first fault is repaired encounters a second type of fault, the second type of fault being the same as the first fault (second fault).

6. The method of claim 5 in which the first type of fault comprises an indication of intrusive repair.

7. The method of claim 5 in which the second type of fault comprises an indication of inadequate diagnostics.

8. The method of claim 1 wherein the intrusivability comprises the probability that a repair action will cause a new fault.

9. The method of claim 8 wherein an intrusivability probability of more than 0.10 is a warning for further analysis of data.

10. The method of claim 1 wherein the one-fixability comprises the probability that a diagnostic process will fix the fault on the first repair attempt.

11. The method of claim 10 wherein a one-fixability probability of less than 0.90 is a warning for further analysis of data.

12. A system for evaluating intrusive repair for a class of devices comprising:

means for providing a threshold time window for analyzing failure data; and

means for utilizing a causal diagram based on the analysis of the failure data for the class of devices to indicate the probability of intrusive repair determined by a measurement of intrusivability and one-fixability.

13. The system of claim 12 in which the failure data is categorized as classes of faults which are identified utilizing error log analysis techniques.

14. The system of claim 13 in which classes of devices comprise computer processing systems.

15. The system of claim 14 in which classes of faults comprise CPU faults, memory faults, power supply faults, software faults and I/O faults.

16. The system of claim 12 in which the causal diagram comprises:

first path wherein a device after a first fault is repaired requires no further repair action;

a second path wherein the device after a first fault is repaired encounters a first type of fault, the first type of fault being different than the first fault (second fault); and

a third path wherein the device after the first fault is repaired encounters a second type of fault, the second type of fault being the same as the first fault (second fault).

17. The system of claim 16 in which the first type of fault comprises an indication of intrusive repair.

18. The system of claim 16 in which the second type of fault comprises an indication of inadequate diagnostics.

19. The system of claim 12 wherein the intrusivability comprises the probability that a repair action will cause a new fault.

20. The system of claim 12 wherein the one-fixability comprises the probability that a diagnostic process will fix the fault on the first repair attempt.

21. The system of claim 19 wherein an intrusivability probability of more than 0.10 is a warning for further analysis of data.

22. The system of claim 20 wherein a one-fixability probability of less than 0.9 is a warning for further analysis of data.