Model based diagnosis and repair for event logs

• US 20060085689A1
• Filed: 09/30/2004
• Published: 04/20/2006
• Est. Priority Date: 09/30/2004
• Status: Active Grant

• Current Assignee: Siemens Healthcare GMBH (Siemens AG)
• Original Assignee: Parker-Hannifin Corporation
• Inventors: Bjorsne, Tord
• Granted Patent: US 7,373,552 B2
• Time in Patent Office: Days
• US Class Current: 714/39
• CPC Class Codes: G06F 11/0748 in a remote unit communicating with a single-box computer node experiencing an error/fault remote testing G06F11/2294
• CPC Class Codes: G06F 11/079 Root cause analysis, i.e. error or fault diagnosis in a hardware test environment G06F11/22; in a software test environment G06F11/36
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06F 11/25 Testing of logic operation, e.g. by logic analysers

Abstract

A system and appertaining method isolates a hardware or user error in a software controlled apparatus e.g., an NMR-apparatus. A diagnostic function is added to an event log that uses a causality model to analyze the event log. A series of events in the event log is evaluated by comparing the events with at least one of a number of root cause related patterns, and a best match of these patterns is used to assign a root cause for the error. Various repair actions may be associated based on a probabilistic model. Additional information may be utilized to modify or improve respective probabilities associated with causes and/or repair actions. This additional information may utilize other automated diagnostic information or may query users for additional information.

Claims

1. A method for isolating a hardware or user error in a software controlled apparatus using recorded events and a causality model, that relates at least one root cause to at least one root cause related pattern of the events, comprising:

recording a series of events;

evaluating the series of events by comparing it with at least one of the root cause related patterns, thereby producing matched patterns;

determining a best match of the matched patterns; and

assigning the root cause that relates to the best matching pattern of the error.

2. The method as claimed in claim 1, wherein one of the root cause related patterns comprises events of the group consisting of error events, information events, and success events.

3. The method as claimed in claim 1, further comprising:

obtaining user-provided additional information by interacting with a user; and

utilizing the user-provided additional information in one of the root cause related patterns.

4. The method as claimed in claim 1, further comprising:

obtaining test-provided additional information by testing the apparatus with test routines; and

utilizing the test-provided additional information in one of the root cause related patterns.

5. The method as claimed in claim 4, wherein one of the root cause related patterns is related with probability factors to more than one root cause.

6. The method as claimed in claim 1, wherein the causality model additionally comprises at least one repair action that is related to at least one of the root causes, the method further comprising providing a chosen repair action to a user after isolating the error.

7. The method as claimed in claim 6, wherein at least one of the repair actions is related with probability factors to more than one root causes.

8. The method as claimed in claim 6, wherein at least one of the root causes is related with probability factors to more than one repair action.

9. The method according to claim 6, further comprising:

automatically executing the chosen repair action.

10. The method according to claim 6, further comprising:

prompting the user to manually execute the chosen repair action.

11. The method as claimed in claim 1, further comprising:

utilizing a Bayesian Network in order to determine a weighting of probability factors.

12. The method as claimed in claim 1, further comprising:

utilizing an event filter to filter events for presentation to a user.

13. A system for isolating a hardware or user error in a software controlled apparatus, comprising:

an event log into which events of the apparatus are recorded;

a table comprising a list of possible root causes for various problems or states;

a diagnosis module configured to implement a causality model relating at least one root cause from the table to at least one root cause related pattern of events by producing matched root cause related patterns to a series of events, determining a best match of the matched patterns, and assigning a root cause that relates to the best matching pattern of the error.

14. The system according to claim 13, further comprising:

a Bayesian Network as a part of the diagnosis module configured to determine a weighting of probability factors.

15. The system according to claim 13, wherein the event log comprises:

an internal event log; and

a user event log;

the system further comprising;

an event filter that filters the internal event log to produce the user event log.

16. The system according to claim 13, further comprising:

a table of repair actions, the diagnosis module being configured to relate at least one of the repair actions to at least one of the root causes.

17. The system according to claim 16, further comprising:

automated repair modules configured to execute the repair actions.