Process pair protection for complex applications

• US 6,622,261 B1
• Filed: 08/01/2002
• Issued: 09/16/2003
• Est. Priority Date: 04/09/1998
• Status: Expired due to Fees

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Compaq Computer Corporation (HP Inc.)
• Inventors: Young, Charles, De Leon, Irma, Jones, Jill A., Sanderson, Stephen M., Gordon, Glen W., Laranjeira, Luiz A., Tien, Yuan Sylvia, Davidson, Thomas J.
• Primary Examiner(s): Iqbal, Nadeem
• Application Number:
  US10/210,904
• Time in Patent Office: 411 Days
• Field of Search: 714/11, 714/12, 714/13, 714/15, 714/25, 714/31, 714/2, 714/6, 714/7, 714/10, 714/47, 709/102, 709/107, 709/220, 709/221
• US Class Current: 714/11
• CPC Class Codes: G06F 11/1482 by means of middleware or OS functionality
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component
• CPC Class Codes: G06F 11/2048 where the redundant components share neither address space nor persistent storage
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups

Abstract

Techniques using process-pair protection of complex applications are disclosed which provide fast and stateful application failover.

Claims

1. An apparatus for providing continuous availability to complex applications -through the use of process-pair protection to allow fast and stateful application failover, the apparatus comprising:

a primary process-pair manager configured to manage a primary instance of a complex application, the primary process-pair manager defining a plurality of states and conditions that trigger transitions between states, each transition associated with one or more actions executed by the primary process-pair manager; and

a backup process-pair manager configured to manage a backup instance of the complex application, the backup process-pair manager and the backup instance of the complex application for replacing the primary process-pair manager and the primary instance of the complex application in the event of failure associated with the primary instance of the complex application.

2. The apparatus of claim 1 wherein the one or more actions executed by the primary process-pair manager include at least one of managing resources, commanding the primary instance of a complex application to change state, checkpointing a new state, and sending a message to the backup process-pair manager.

3. The apparatus of claim 1 wherein backup process-pair manager defines a plurality of states and conditions that trigger transitions between states, each transition associated with one or more actions executed by the backup process-pair manager.

4. The apparatus of claim 3 wherein the one or more actions executed by the backup process-pair manager include at least one of managing resources, commanding the backup instance of a complex application to change state, checkpointing a new state, and sending a message to the primary process-pair manager.

5. The apparatus of claim 1 wherein a watchdog process monitors the primary process-pair manager and restarts the primary process-pair manager in case of failure.

6. The apparatus of claim 1 wherein a watchdog process monitors the backup process-pair manager and restarts the backup process-pair manager in case of failure.

7. The apparatus of claim 1 further comprising:

a node watcher configured to inform the backup process-pair manager of failures associated with the primary process-pair manager.

8. The apparatus of claim 1 further comprising:

a node watcher configured to inform the primary process-pair manager of failures associated with the backup process-pair manager.

9. The apparatus of claim 1 wherein the primary instance of the complex application is adapted to periodically perform a checkpointing operation by sending internal state information to the backup instance of the complex application.

10. The apparatus of claim 1 further comprising:

at least one application administration module configured to provide an interface between one of the process-pair managers and components of the corresponding complex application.

11. The apparatus of claim 1 further comprising:

at least one interapplication communication module configured to facilitate communication between the process-pair managers.

12. The apparatus of claim 1 wherein the apparatus is a computer program product stored on a computer readable medium.

13. The apparatus of claim 1 wherein the primary process-pair manager is located on a primary computer system and the backup process-pair manager is located on a backup computer system.

14. A method for providing high availability to complex applications through the use of process-pair protection to allow fast and stateful application failover, the method comprising:

receiving an indication of a failure associated with a primary instance of an application that includes a plurality of processes;

in response to the indicated failure, transitioning a primary process-pair manager to a disabled state thereby causing the primary instance of the application to stop providing service; and

transitioning a backup process-pair manager from a backup state to a primary state thereby causing an initialized backup instance of the application to become the primary service provider by acquiring state information that was checkpointed by the primary application.

15. The method of claim 14 wherein the transitioning the primary process-pair manager to a disabled state includes sending a message thereby commanding each process of the backup instance of the application to become primary.

16. The method of claim 14 wherein prior to receiving the indication of the failure, the method further comprises:

periodically checkpointing state information associated with the processes of the primary instance to the backup instance of the application.

17. The method of claim 14 wherein transitioning the primary process-pair manager to a disabled state further causes a cleanup operation where processes of the primary instance of the application that survived the failure are stopped.

18. The method of claim 14 wherein transitioning the primary process-pair manager to a disabled state further causes a transfer of resources that need to be used by the backup instance of the application as it becomes primary.

19. The method of claim 14 wherein the disabled state is one of down, degraded, or failed.

20. An apparatus for providing continuous availability to complex applications through the use of process-pair protection to allow fast and stateful application failover, the apparatus comprising:

a primary process-pair manager means configured to manage a primary instance of a complex application, the primary process-pair manager means defining a plurality of states and conditions that trigger transitions between states, each transition associated with one or more actions executed by the primary process-pair manager; and

a backup process-pair manager means configured to manage a backup instance of the complex application, the backup process-pair manager means and the backup instance of the complex application for replacing the primary process-pair manager means and the primary instance of the complex application in the event of failure associated with the primary instance of the complex application.

21. The apparatus of claim 20 wherein a watchdog process means monitors the primary process-pair manager means and restarts the primary process-pair manager means in case of failure.

22. The apparatus of claim 20 wherein a watchdog process means monitors the backup process-pair manager means and restarts the backup process-pair manager means in case of failure.

23. The apparatus of claim 20 further comprising:

a node watcher means configured to inform the backup process-pair manager means of failures associated with the primary process-pair manager means.

24. The apparatus of claim 20 further comprising:

a node watcher means configured to inform the primary process-pair manager means of failures associated with the backup process-pair manager means.

25. The apparatus of claim 20 further comprising:

at least one application administration means configured to provide an interface between one of the process-pair manager means and components of the corresponding complex application.

26. The apparatus of claim 20 further comprising:

at least one interapplication communication means configured to facilitate communication between the process-pair manager means.

27. The apparatus of claim 20 wherein the primary process-pair manager means is located on a primary system means and the backup process-pair manager means is located on a backup system means.

28. A computer program product, stored on electronically-readable medium, for providing high availability to complex applications through the use of process-pair protection to allow fast and stateful application failover, wherein in response to the computer program product being executed by a processor, the processor performs the steps of:

receiving an indication of a failure associated with a primary instance of an application that includes a plurality of processes;

in response to the indicated failure, transitioning a primary process-pair manager to a disabled state thereby causing the primary instance of the application to stop providing service; and

transitioning a backup process-pair manager from a backup state to a primary state thereby causing an initialized backup instance of the application to become the primary service provider by acquiring state information that was checkpointed by the primary application.

29. The computer program product of claim 28 wherein the transitioning the primary process-pair manager to a disabled state causes a message to be sent thereby commanding each process of the backup instance of the application to become primary.

30. The computer program product of claim 28 wherein transitioning the primary process-pair manager to a disabled state further causes a cleanup operation where processes of the primary instance of the application that survived the failure are stopped.

31. The computer program product of claim 28 wherein transitioning the primary process-pair manager to a disabled state further causes a transfer of resources that need to be used by the backup instance of the application as it becomes primary.

32. The computer program product of claim 28 wherein prior to receiving the indication of the failure, the processor performs the step of:

periodically checkpointing state information associated with the processes of the primary instance to the backup instance of the application.