System and method for achieving autonomic computing self-healing, utilizing meta level reflection and reasoning
First Claim
Patent Images
1. A method of handling errors, said method comprising:
- under control of a base level,detecting an error in a production environment;
providing a reification message comprising data about said error to a rule-based expert system in a meta level, wherein the rule-based expert system in said meta level receives the reification message;
analyzes said data, using knowledge of base-level computational components;
chooses a self-healing action for said error; and
returns a reversion message comprising a signal to implement the self-healing action; and
responsive to receiving the signal at the base level, implementing said self-healing action at said base level,wherein the error is an out of memory condition; and
wherein the production environment comprises a heap and wherein the self-healing action comprises at least one of reconstructing objects in the heap or storing objects from the heap that are not accessed frequently off-line.
2 Assignments
0 Petitions
Accused Products
Abstract
A mechanism is provided for achieving autonomic computing self-healing utilizing meta level reflection. In a base level, a monitor detects an error in a production environment. The base level provides a reification message comprising data about the error to a meta level. A reasoning system in the meta level receives the reification message and analyzes the data using knowledge of computational components in the base level. The reasoning system identifies a self-healing action for the error and returns a reversion message comprising a signal to implement the self-healing action. Responsive to receiving the signal, the base level implements the self-healing action.
88 Citations
16 Claims
-
1. A method of handling errors, said method comprising:
- under control of a base level,
detecting an error in a production environment; providing a reification message comprising data about said error to a rule-based expert system in a meta level, wherein the rule-based expert system in said meta level receives the reification message;
analyzes said data, using knowledge of base-level computational components;
chooses a self-healing action for said error; and
returns a reversion message comprising a signal to implement the self-healing action; andresponsive to receiving the signal at the base level, implementing said self-healing action at said base level, wherein the error is an out of memory condition; and wherein the production environment comprises a heap and wherein the self-healing action comprises at least one of reconstructing objects in the heap or storing objects from the heap that are not accessed frequently off-line. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
- under control of a base level,
-
10. A system of handling errors, said system comprising:
-
means for detecting an error in a production environment; means for providing a reification message comprising data about the error to a rule-based expert system in a meta level; means for performing introspection at said rule-based expert system, in response to said reification message, wherein the introspection uses knowledge of computational components of the production environment to identify a self-healing action for the error; means for providing a reversion message from the rule-based expert system to the production environment, wherein the reversion message comprises a signal to implement the self-healing action; and means for implementing the self-healing action at said production environment, wherein the error is an out of memory condition; and wherein the production environment comprises a heap and wherein the self-healing action comprises at least one of reconstructing objects in the heap or storing objects from the heap that are not accessed frequently off-line. - View Dependent Claims (11, 12, 13, 14, 15, 16)
-
Specification