SYSTEM AND METHOD FOR SELF-HEALING

• US 20120124410A1
• Filed: 08/22/2011
• Published: 05/17/2012
• Est. Priority Date: 11/16/2010
• Status: Active Grant

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Chun, In Geol, Park, Seung Min, Kim, Won Tae
• Granted Patent: US 9,372,742 B2
• Time in Patent Office: Days
• US Class Current: 714/2
• CPC Class Codes: G06F 11/0736 in functional embedded systems, i.e. in a data processing system designed as a combination of hardware and software dedicated to performing a certain function testing or monitoring of automated control systems G05B23/02
• 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

Abstract

Provided are a system and a method for self-healing in a critical system. The present invention monitors a current situation of the critical system, determines whether a system has an error by analyzing the monitoring result, judges whether to perform self-healing in a current state or drive safety software which provides a minimum basic service according to self-healing of the system error or not when the system error occurs, and evaluates self-healing performance after healing the system error. According to exemplary embodiments of the present invention, it is possible to continuously provide a software service and further improve the reliability of the self-healing system through the evaluation of the self-healing performance.

Claims

1. A self-healing system, comprising:

a system error determining unit analyzing a monitoring result of a system environment or components constituting the system and determining whether the system error occurs on the basis of the analyzing result;

a countermeasure selecting unit selecting a countermeasure depending on whether or not to self-heal the system error when the system error occurs;

a software service providing unit continuously providing the software service in the system by reflecting the selected countermeasure; and

a healing performance evaluating unit evaluating healing performance on the basis of a healing result when the selected countermeasure is self-healing of the system error.

2. The self-healing system according to claim 1, wherein the countermeasure selecting unit selects self-healing software having the error or executing safety software associated with the software having the error as the countermeasure, or

the software service providing unit includes:

a self-healing portion self-healing the software having the error when the selected countermeasure is self-healing of the system error; and

a safety software executing unit executing the safety software associated with the software having the error when the selected countermeasure is not the self-healing of the system error

3. The self-healing system according to claim 2, wherein the self-healing portion includes:

an internal healing portion internally healing the software by controlling a healing code in the software having the error; and

an external healing portion performing healing by adding/modifying/deleting contents of the corresponding software outside the software having the error.

4. The self-healing system according to claim 2, wherein the software service providing unit further includes:

a service executing portion managing an actuator set for each component and connects a component requesting execution of software and the execution requested software with each other by using the actuator; and

a service operation controlling portion controlling the self-healing portion or the safety software executing portion according to the selected countermeasure and simultaneously executing software associated with the same service by considering the system environment.

5. The self-healing system according to claim 1, further comprising a monitoring unit monitoring an environment of at least one critical system or components constituting at least one critical system.

6. The self-healing system according to claim 5, further comprising a defining unit defining at least one of information on a class configuring the component, state information of the component, and relationship information between the components.

7. The self-healing system according to claim 6, wherein the monitoring unit performs monitoring on the basis of the defined information, and

the healing performance evaluating unit evaluates a system characteristic including at least one of reliability and system performability by the defined information as the healing performance.

8. The self-healing system according to claim 1, wherein the healing performance evaluating unit includes:

a policy database storing a monitoring policy which forms the basis of a monitoring method, a performance evaluation policy which forms the basis of evaluation of healing performance, and a self-healing policy associated with self-healing;

a performance evaluation model analyzing portion separating and analyzing a model for performance evaluation from models to be monitored and determining the performance evaluation policy on the basis of the analysis result; and

a healing performance evaluating engine evaluating the self-healing performance on the basis of the determined performance evaluation policy.

9. The self-healing system according to claim 8, wherein the healing performance evaluating engine includes:

an information collecting portion collecting information associated with the system environment after self-healing;

a policy loading portion loading the determined performance evaluation policy when the performance evaluation policy is determined by the performance evaluation model analyzing portion on the basis of the collected information; and

a performance evaluating portion quantitatively evaluating the self-healing performance on the basis of the loaded performance evaluation policy, and

the performance evaluating includes;

an analysis information generating portion analyzing the performance evaluation model for evaluating self-healing performance on the basis of information collected by an information collecting portion and the policy stored in the policy database and generating analysis information on the self-healing performance by comparing a state of the performance evaluation model before self-healing and a state of the performance evaluation model after self-healing with each other; and

a policy determining portion determining the performance evaluation policy on the basis of the generated analysis information.

10. The self-healing system according to claim 1, further comprising a healing result learning unit that learns the self-healing result for the self-healed software on the basis of the evaluated healing performance and performs learning on the basis of a mechanical learning algorithm when the system performance is deteriorated after self-healing as compared with before self-healing.

11. The self-healing system according to claim 10, wherein the healing result learning unit modifies even a predetermined stored monitoring policy according to the performed learning.

12. A self-healing method, comprising:

system error determining of analyzing a monitoring result of a system environment or components constituting the system and determining whether the system error occurs on the basis of the analyzing result;

countermeasure selecting of selecting a countermeasure depending on whether or not to self-heal the system error when the system error occurs;

software service providing of continuously providing the software service in the system by reflecting the selected countermeasure; and

healing performance evaluating of evaluating healing performance on the basis of a healing result when the selected countermeasure is self-healing of the system error.

13. The self-healing method of claim 12, wherein in the countermeasure selecting, self-healing software having the error or executing safety software associated with the software having the error is selected as the countermeasure, or

the software service providing includes:

any one of

self-healing of self-healing the software having the error when the selected countermeasure is self-healing of the system error; and

safety software executing of executing the safety software associated with the software having the error when the selected countermeasure is not the self-healing of the system error.

14. The self-healing method according to claim 13, wherein the self-healing includes:

internal healing of internally healing the software by controlling a healing code in the software having the error; and

external healing of performing healing by adding/modifying/deleting contents of the corresponding software outside the software having the error.

15. The self-healing method according to claim 13, wherein the software service providing further includes:

service executing of managing an actuator set for each component and connects a component requesting execution of software and the execution requested software with each other by using the actuator; and

service operation controlling of controlling the self- healing portion or the safety software executing portion according to the selected countermeasure and simultaneously executing software associated with the same service by considering the system environment.

16. The self-healing method according to claim 12, further comprising monitoring an environment of at least one critical system or components constituting at least one critical system.

17. The self-healing method according to claim 16, further comprising defining at least one of information on a class configuring the component, state information of the component, and relationship information between the components.

18. The self-healing method according to claim 12, wherein the healing performance evaluating includes:

policy database constructing of storing a monitoring policy which forms the basis of a monitoring method, a performance evaluation policy which forms the basis of evaluation of healing performance, and a self-healing policy associated with self-healing;

performance evaluation model analyzing of separating and analyzing a model for performance evaluation from models to be monitored and determining the performance evaluation policy on the basis of the analysis result; and

healing performance evaluating engine driving of evaluating the self-healing performance on the basis of the determined performance evaluation policy.

19. The self-healing method according to claim 18, wherein the healing performance evaluating engine driving includes:

information collecting of collecting information associated with the system environment after self-healing;

policy loading of loading the determined performance evaluation policy when the performance evaluation policy is determined by the performance evaluation model analyzing portion on the basis of the collected information; and

performance evaluating of quantitatively evaluating the self-healing performance on the basis of the loaded performance evaluation policy, and

the performance evaluating model analyzing includes;

analysis information generating of analyzing the performance evaluation model for evaluating self-healing performance on the basis of information collected by an information collecting portion and the policy stored in the policy database and generating analysis information on the self-healing performance by comparing a state of the performance evaluation model before self-healing and a state of the performance evaluation model after self-healing with each other; and

policy determining of determining the performance evaluation policy on the basis of the generated analysis information.

20. The self-healing method according to claim 12, further comprising healing result learning of learning the self-healing result for the self-healed software on the basis of the evaluated healing performance and performing learning on the basis of a mechanical learning algorithm when the system performance is deteriorated after self-healing as compared with before self-healing.