Method and apparatus for enabling dynamic self-healing of multi-media services
View Patent ↗In the service architecture of the invention, each service is viewed as a hierarchical arrangement of service components. At least one management agent is associated with each layer, and is used to manage and preserve the expected performance of the service components. The management agents cooperatively exchange management information regarding the service components at their layers to permit service self-management and self-healing, enabling restoration and recovery with minimum or no human intervention.
1. A system comprising a service architecture comprising:
a hierarchy of service layers including a service user layer comprising an interface by which a user builds a service instance, a service access layer for controlling access to service instances by the user, a service capability layer comprising service capability instances associated with the service instances and a service element layer comprising elements available to the service capability instances, wherein each service instance is a self-managed aggregation of service capabilities delivered to the user via assigned service components according to a service level agreement; and
a plurality of hierarchically linked management agents, at least one management agent associated with each one of the service layers, for cooperatively managing, with other management agents associated with a service layer, a plurality of service layer components of the service instance at and below the associated service layer including discovering, monitoring and, in the event of detection of a degradation of performance of the service instance, replacing service layer components associated with a service instance at and below the associated service layer.
2. The service architecture of claim 1 , further comprising:
an information repository apportioned into a plurality of management repositories, wherein each management repository is associated with one of the management agents, and accessible by the associated management agent and any management agents senior to the associated management agent in the hierarchy of management agents.
3. The service architecture of claim 2 , wherein information repository further comprises an information records component for storing information related to service component behavior.
4. The service architecture of claim 3 wherein the information includes an event log associated with the service components.
5. The service architecture of claim 3 wherein the information includes performance information.
6. The service architecture of claim 3 wherein the information includes accounting information.
7. The service architecture of claim 3 , wherein the information repository further comprises an information model component for storing object information associated with the service components in the hierarchy.
8. The service architecture of claim 7 wherein the object information includes physical topology data associated with the service component.
9. The service architecture of claim 7 wherein the object information includes logical topology data associated with the service components.
10. The service architecture of claim 7 wherein the object information includes a model of the hierarchy of the service components.
11. The service architecture of claim 7 wherein the object information includes service component objects.
12. The service architecture of claim 3 wherein the information repository further comprises an operations model component comprising routines capable of being invoked to access information records and information models.
13. The service architecture of claim 12 wherein the operations models include a set of recovery operation service instances that are indexed using a cause-condition pair.
14. The service architecture of claim 12 wherein the operation models include a set of workflow management routines.
15. The service architecture of claim 12 wherein the operation models include a set of performance operations.
16. The service architecture of claim 1 , wherein each of the management agents interact with an event-log, for logging events related to the service layer components in the associated service layer.
17. The service architecture of claim 16 , wherein each of the management agents includes a learning module, coupled to the event log, for controlling operation of the service component and a list of alternative service components to be used to replace the service component upon detection of a degradation of service component performance.
18. The service architecture of claim 1 wherein the service element layer is at a first lowest layer in the hierarchy and includes service elements and network elements.
19. The service architecture of claim 18 , wherein the service capability layer is at a second layer in the hierarchy that is a parent layer to the first layer.
20. The service architecture of claim 19 , wherein the service access layer is at a third layer in the hierarchy above the second layer.
21. The service architecture of claim 1 wherein the management agents in the hierarchy communicate by forwarding request and notification messages up and down the hierarchy.
22. A computer-implemented method for dynamic self healing of a service comprising a plurality of hierarchically ordered service component layers comprises the steps of:
deploying a plurality of cooperating management agents including a management agent for each service component layer;
discovering, by at least one management agent associated with at least one service layer, a service component associated with the service component layer;
monitoring, by the management agent associated with the service component layer, a performance of the service component to determine if the performance is in accordance with a contracted service level agreement of the service instance, wherein the service instance is a self-managed aggregation of service capabilities associated with service components, and is delivered to the user according to the service level agreement;
detecting, by the management agent, a performance issue associated with the service component;
upon detecting the performance issue, the management agent forwarding an alarm to a parent management agent associated with a parent service component of the service component; and
the child management agent triggering recovery at the service component, including discovering an alternate service component to replace the service component to allow continued support of the service, selecting, in cooperation with peer management agents, the alternate service component wherein the plurality of hierarchically ordered service component layers includes a service user layer comprising an interface by which a user builds a service instance, a service access layer for controlling access to service instances by the user, a service capability layer comprising service capability instances associated with the service instances and a service element layer comprising elements available to the service capability instances.
23. The method of claim 22 including the steps of logging, at each of the service component layers, a handling of the performance issue.
24. The method of claim 23 including the step of reviewing a log of past handling of performance issues and a list of alternate service components to proactively select an appropriate recovery action including replacing the service component with the alternate service component for supporting the service upon detection of a degradation of performance of the service component.
25. The method of claim 22 wherein the service element layer is at a first lowest layer in the hierarchy and includes service elements and network elements.
26. The method of claim 25 , wherein the service capability layer is at a second layer in the hierarchy that is a parent layer to the first layer.
27. The method of claim 26 , wherein the service access layer is at a third layer in the hierarchy above the second layer.
28. A computer-implemented method for building a self-managed service network, wherein the service architecture comprises hierarchically linked service component layers includes the steps of:
deploying a plurality of management agents in the network, at least one management agent associated with each service component layer for cooperatively managing, with other management agents associated with the service component layer, a plurality of service layer components at and below the associated service layer, the service component layers including a service user layer comprising an interface by which a user builds a service instance, a service access layer for controlling access to service instances by the user, a service capability layer comprising service capability instances associated with the service instances and a service element layer comprising elements available for use by the service capability instances, wherein each service instance is a self-managed aggregation of service capabilities delivered to the user according to a service level agreement and wherein each management agent includes functionality for managing service components at and below the associated layer of the network;
a management agent associated with at least one service component layer discovering a service component associated with the at least one service component layer; and
registering service components in the network, the service component being registered at a particular layer of the network, including storing information associated with the service component in a repository accessible by the management agent associated with the particular layer to allow service components to be selectively replaced by the management agent corresponding to the particular layers in response to a detection of a degradation of performance of the service instance.