Detecting events of interest for managing components on a high availability framework

• US 7,350,098 B2
• Filed: 12/02/2002
• Issued: 03/25/2008
• Est. Priority Date: 11/30/2001
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Brower, David, Colrain, Carol Lyndall
• Primary Examiner(s): Beausoliel; Robert
• Assistant Examiner(s): GUYTON, PHILIP A
• Application Number:
  US10/308,866
• Publication Number:
  US 20030105993A1
• Time in Patent Office: 1,940 Days
• Field of Search: 714/4, 714/43, 709/224
• US Class Current: 714/4.3
• CPC Class Codes: G06F 2209/5011 Pool
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

Events of interest are detected in order to manage a high availability framework. In a framework in which a plurality of components are executing, the components are periodically polled to detect occurrence of the event of interest. A monitor is also established for one or more of the components. After the first component causes the event of interest to occur, the monitor communicating the event of interest to the framework without waiting for the framework to poll the first component.

Claims

1. A method of detecting an event of interest, the method comprising the steps of:

in a framework in which a plurality of members are executing, establishing a monitor for at least a first member in the plurality of members to detect occurrence of the event of interest, the first member residing on a first node, wherein each of the plurality of members performs a service;

after the first member causes the event of interest to occur, the monitor communicating the event of interest to the framework; and

identifying at least an active second member to replace the first member in response to the monitor communicating the event of interest, the second member being configured to perform a service that is comparable to the service performed by the first member.

2. The method of claim 1, further comprising periodically polling the members to detect occurrence of the event of interest, and wherein the step of establishing a monitor is performed without waiting for the framework to poll the first member.

3. The method of claim 1, further comprising the monitor detecting that the event of interest is caused by the first member.

4. The method of claim 1, wherein the monitor communicating the event of interest to the framework includes the monitor communicating that the first member has failed.

5. The method of claim 1, further comprising establishing, on a plurality of nodes, a composite resource that includes the plurality of members, wherein each member of said plurality of said members is capable of providing a comparable service, wherein the first member is one of the plurality of members of the composite resource, and wherein the composite resource is at least partially defined by recovery initiation of the composite resource occurring when each member of the plurality of members has ceased to be active.

6. The method of claim 5, further comprising establishing a monitor for each of the plurality of members of the composite resource, and initiating recovery of the composite resource when the monitor of each of the plurality of members detects that the member has failed.

7. The method of claim 1, further comprising the steps of:

identifying a first resource that the service provided by the first member depends on; and

in response to the event of interest, triggering a second resource that is comparable to the first resource on which the service provided by the second member depends.

8. The method of claim 1, wherein the monitor communicating the event of interest is configured to communicate that the first member is starting.

9. The method of claim 1, wherein the monitor communicating the event of interest to the framework includes the monitor communicating that a service previously provided by the first member is to be provided by another member of the composite resource.

10. The method of claim 1, wherein the monitor communicating the event of interest to the framework includes the monitor communicating that the first member has stopped and not failed.

11. The method of claim 1, wherein the monitor communicating the event of interest to the framework includes the monitor communicating that the first member has stopped and failed.

12. The method of claim 1, wherein the monitor communicating the event of interest to the framework includes the monitor communicating that the first node has failed.

13. The method of claim 1, further comprising:

communicating that the first member needs to be checked.

14. The method of claim 2, wherein the step of periodically polling said members to detect occurrence of the event of interest is at least temporarily stopped in response to the step of the monitor communicating the event of interest to the framework.

15. A method of detecting a member failure, the method comprising the steps of:

on a framework, establishing a composite resource that includes a plurality of members, where each member of said plurality of said members is capable of providing a comparable service;

providing said service to a component that requests said service by;

arranging for said service to be provided to said component by a first member of said composite resource;

establishing a monitor for at least the first member;

the monitor communicating to the framework that the first member has ceased to be active; and

automatically causing the service to be provided to said component by another member of said plurality of members; and

maintaining a state of the composite resource independently of the state of each member of said composite resource.

16. The method of claim 15, further comprising periodically polling the first member to determine occurrence if the first member ceases to be active.

17. The method of claim 15, wherein the monitor communicating to the framework when the first member ceases to be active includes the monitor detecting that a first node where the first member resides has failed.

18. The method of claim 17, wherein automatically causing the service to be provided to said component by another member of said plurality of members includes automatically causing the service to be provided by a second member in the plurality of members, wherein the second member resides on a second node that is active.

19. The method of claim 18, wherein automatically causing the service to be provided by a second member in the plurality of members includes executing code on the first framework, execution of the code causing the service to be provided by the second member, wherein the framework is shared by the first member and the second member.

20. The method of claim 15, further comprising the steps of:

identifying a separate resource from the composite resource, wherein the composite resource is dependent on the comparable service and wherein prior to the first member ceasing to be active, the component receives the comparable service from the first member; and

providing the service to the separate resource from the second member.

21. The method of claim 20, wherein providing the service to the separate resource from the second member includes starting a corresponding separate resource on the second node that is comparable to the separate resource from the first member on the first node.

22. The method of claim 18, further comprising:

identifying a first resource that the first member is dependent on, wherein prior to the first member ceasing to be active, the first resource resides on the first node and triggers the comparable service from the first member; and

in response to the monitor communicating to the framework that the first member ceases to provide the comparable service, stopping the first resource from executing on the first node; and

starting a second resource on the second node to trigger the comparable service from the second member, wherein the second resource is comparable to the first resource.

23. The method of claim 22, wherein the method further comprises establishing a second monitor for the second member, and wherein the step of starting a second resource on the second node includes detecting that the second member is active using the second monitor.

24. The method of claim 22, wherein the first node is in a first framework and the second node is in a second framework, and wherein starting a second resource on the second node includes communicating between the first framework and the second framework over a network.

25. The method of claim 15, wherein establishing a monitor for at least the first member includes establishing one or more monitors for each member of the composite resource, the one or more monitors determining whether each of the plurality of members are capable of providing the comparable service.

26. The method of claim 25, wherein establishing one or more monitors for each member of the composite resource includes determining whether any of the plurality of members are operational to provide the comparable service in response to any of the plurality of members ceasing to provide the comparable service.

27. The method of claim 25, further comprising the step of initiating recovery of the composite resource upon the one or more monitors determining that each member of said plurality of members has ceased to be active.

28. The method of claim 24, wherein upon the one or more monitors determining that each member of said plurality of members has ceased to be active, the method further comprises reevaluating the composite resource.

29. A computer-readable storage medium for carrying one or more sequences of instructions for detecting an event of interest, wherein execution of the one or more sequences of instructions by one or more processors causes the one or more processors to perform the steps of:

in a framework in which a plurality of members are executing, establishing a monitor for at least a first member in the plurality of members to detect occurrence of the event of interest, the first member residing on a first node, wherein each of the plurality of members performs a service;

after the first member causes the event of interest to occur, the monitor communicating the event of interest to the framework; and

identifying at least an active second member to replace the first member in response to the monitor communicating the event of interest, the second member being configured to perform a service that is comparable to the service performed by the first member.

30. The computer-readable medium of claim 29, further comprising instructions for periodically polling the members to detect occurrence of the event of interest, and wherein the step of establishing a monitor is performed without waiting for the framework to poll the first member.

31. The computer-readable storage medium of claim 29, further comprising instructions for the monitor detecting that the event of interest is caused by the first member.

32. The computer-readable storage medium of claim 29, wherein instructions for the monitor communicating the event of interest to the framework include instructions for the monitor communicating that the first member has failed.

33. The computer-readable storage medium of claim 29, further comprising instructions for performing the step of establishing, on a plurality of nodes, a composite resource that includes the plurality of members, wherein each member of said plurality of said members is capable of providing a comparable service, wherein the first member is one of the plurality of members of the composite resource, and wherein the composite resource is at least partially defined by recovery initiation of the composite resource occurring when each member of the plurality of members has ceased to be active.

34. The computer-readable storage medium of claim 33, further comprising instructions for performing the step of establishing a monitor for each of the plurality of members of the composite resource, and instructions for performing the step of initiating recovery of the composite resource when the monitor of each of the plurality of members detects that the member has, failed.

35. The computer-readable storage medium of claim 29, further comprising instructions for performing the steps of:

identifying a first resource that the service provided by the first member depends on; and

in response to the event of interest, triggering a second resource that is comparable to the first resource on which the service provided by the second member depends.

36. The computer-readable storage medium of claim 29, wherein instructions for the monitor communicating the event of interest are configured to communicate that the first member is starting.

37. The computer-readable storage medium of claim 29, wherein instructions for the monitor communicating the event of interest to the framework include instructions for the monitor communicating that a service previously provided by the first member is to be provided by another member of the composite resource.

38. The computer-readable storage medium of claim 29, wherein instructions for the monitor communicating the event of interest to the framework include instructions for the monitor communicating that the first member has stopped and not failed.

39. The computer-readable storage medium of claim 29, wherein instructions for the monitor communicating the event of interest to the framework include instructions for the monitor communicating that the first member has stopped and failed.

40. The computer-readable storage medium of claim 29, wherein instructions for the monitor communicating the event of interest to the framework include instructions for the monitor communicating that the first node has failed.

41. The computer-readable storage medium of claim 29, further comprising communicating the first member needs to be checked.

42. The computer-readable storage medium of claim 30, wherein instructions for performing the step of periodically polling said members to detect occurrence of the event of interest is at least temporarily stopped in response to the step of the monitor communicating the event of interest to the framework.