Method for mutual computer process monitoring and restart

• US 6,928,585 B2
• Filed: 05/24/2001
• Issued: 08/09/2005
• Est. Priority Date: 05/24/2001
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Bartley, Timothy Simon
• Primary Examiner(s): Bonzo, Bryce P.
• Assistant Examiner(s): MANOSKEY, JOSEPH D
• Application Number:
  US09/864,118
• Publication Number:
  US 20020184295A1
• Time in Patent Office: 1,538 Days
• Field of Search: 714/23, 714/25, 714/38, 714/48
• US Class Current: 714/23
• CPC Class Codes: G06F 11/0715 in a system implementing multitasking multitasking per se G06F9/46
• CPC Class Codes: H04L 41/0893 Assignment of logical groups to network elements
• CPC Class Codes: H04L 41/0894 Policy-based network configuration management

Abstract

The present invention provides a method for detecting a termination of a process within a plurality of processes in a data processing system. A monitoring policy is established, within the plurality of processes, wherein the monitoring policy assigns a first process within the plurality of processes to monitor a second process within the plurality of processes. Responsive to a termination of execution of the second process, a cause of the execution termination is determined by the first process. Responsive to a determination that the second process terminated execution in an abnormal manner, the first process attempts to restart the second process. Furthermore, the present invention provides a method for inserting a process within a plurality of processes containing a first process and a monitoring policy in a data processing system. A request is received from a second process to join the plurality of processes. Responsive to the second process joining the plurality of processes, the first process within the plurality of processes is selected to monitor the second process. The monitoring policy is modified, wherein the monitoring policy assigns the selected first process to monitor the second process for termination of execution.

Claims

1. A method of mutual process monitoring for detecting a termination of a process within a plurality of processes in a data processing system, the method comprising the steps of:

establishing a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system;

responsive to a termination of execution of a monitored process, determining the cause of the execution termination by the monitoring process; and

responsive to a determination that the monitored process terminated execution in an abnormal manner, attempting to restart the monitored process by the monitoring process.

2. The method as recited in claim 1, further comprising:

opening a communications link between the monitoring process and the monitored process; and

posting a blocking read by the monitoring process in order to detect termination of the monitored process.

3. The method as recited in claim 2, wherein the communications link is a watch FIFO (first-in first-out) communications link.

4. The method as recited in claim 3, wherein posting a blocking read to the watch FIFO communications link is performed in a single thread of execution.

5. The method as recited in claim 2, wherein the monitoring process posts a blocking operation on the communications link that completes the blocking operation on termination of the monitored process.

6. The method as recited in claim 2, wherein the blocking operation is performed in a separate execution thread within the monitoring process in order to allow normal process operation of the monitored process to continue.

7. The method as recited in claim 1, further comprising:

responsive to a determination that the monitored process terminated execution in a normal manner, modifying the monitoring policy, wherein the monitoring policy is modified to assign the monitoring process to monitor a different process within the plurality of processes; and

monitoring of the different process by the monitoring process for termination of execution.

8. A method for inserting an additional process within a plurality of processes containing a mutually monitoring process policy in a data processing system, the method comprising the steps of:

establishing a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby fanning a ringof mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system;

receiving a request for an additional process to join the plurality of processes;

responsive to the additional process joining the plurality of processes, randomly selecting a process within the plurality of processes to monitor the additional process; and

modifying the mutually monitoring process policy, wherein the selected process monitors the additional process for termination of execution.

9. The method as recited in claim 8, wherein the additional process monitors the process previously monitored by the selected process.

10. The method as recited in claim 8, wherein the additional process joining the plurality of processes joins as a child of the selected process.

11. The method as recited in claim 8, wherein the request is received by a inter-process communications (IPC) mechanism.

12. The method as recited in claim 8, further comprising:

locking a state file in a join communications link; and

blocking any other processes attempting to simultaneously join the plurality of processes.

13. The method as recited in claim 12, wherein the communications link is a first-in first-out communications link.

14. The method as recited in claim 12, wherein blocking any other processes attempting to simultaneously join the plurality of processes is by a file lock.

15. A data processing system for mutual process monitoring for detecting a termination of a process within a plurality of processes in a data processing system, comprising:

establishing means for establishing a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system;

determining means, responsive to a termination of execution of a monitored process, for determining the cause of the execution termination by the monitoring process; and

attempting means, responsive to a determination that the monitored process terminated execution in an abnormal manner, for attempting to restart the monitored process by the monitoring process.

16. The data processing system as recited in claim 15, further comprising:

opening means for opening a communications link between the monitoring process and the monitored process; and

posting means for posting a blocking read by the monitoring process in order to detect termination of the monitored process.

17. The data processing system as recited in claim 16, wherein the communications link is a watch FIFO (first-in first-out) communications link.

18. The data processing system as recited in claim 17, wherein posting a blocking read to the watch FIFO communications link is performed in a single thread of execution.

19. The data processing system as recited in claim 16, wherein the monitoring process posts a blocking operation on the communications link that completes the blocking operation on termination of the monitored process.

20. The data processing system as recited in claim 16, wherein the blocking operation is performed in a separate execution thread within the monitoring process in order to allow normal process operation of the monitored process to continue.

21. The data processing system as recited in claim 15, further comprising:

modifying means, responsive to a determination that the monitored process terminated execution in a normal manner, for modifying the monitoring policy, wherein the monitoring policy is modified to assign the monitoring process to

monitor a different process within the plurality of processes; and

monitoring means for monitoring of the different process by the monitoring process for termination of execution.

22. A data processing system for inserting an additional process within a plurality of processes containing a mutually monitoring process policy in a data processing system, the method comprising the steps of:

establishing means for establishing a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system;

receiving means for receiving a request for an additional process to join the plurality of processes;

selecting means, responsive to the additional process joining the plurality of processes, for randomly selecting a process within the plurality of processes to monitor the additional process; and

modifying means for modifying the mutually monitoring process policy, wherein the selected process monitors the additional process for termination of execution.

23. The data processing system as recited in claim 22, wherein the additional process monitors the process previously monitored by the selected process.

24. The data processing system as recited in claim 22, wherein the additional process joining the plurality of processes joins as a child of the selected process.

25. The data processing system as recited in claim 22, wherein the request is received by a inter-process communications (IPC) mechanism.

26. The data processing system as recited in claim 22, further comprising:

locking means for locking a file used to join the communications link; and

blocking means for blocking any other processes attempting to simultaneously join the plurality of processes.

27. The data processing system as recited in claim 26, wherein the communications link is a FIFO communications link.

28. The data processing system as recited in claim 26, wherein blocking any other processes attempting to simultaneously the plurality of processes is by a file lock.

29. A data processing system, comprising:

a bus system;

a memory, containing a set of instructions and a plurality of processes, connected to the bus system; and

a processing unit connected to the bus system, wherein the processing unit executes the set of instructions from the memory to establish a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system, responsive to a termination of execution of a monitored process, the processing unit instructs the monitoring process to determine the cause of the execution termination, and responsive to a determination that the monitored process terminated execution in an abnormal manner, the processing unit instructs the monitoring process to attempt to restart the monitored process.

30. A data processing system, comprising:

a bus system;

a memory, containing a set of instructions and a plurality of processes, connected to the bus system; and

a processing unit connected to the bus system, wherein the processing unit executes the set of instructions from the memory to establish a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system, receive a request for an additional process to join the plurality of processes, respond to the additional process joining the plurality of processes, the processing unit randomly selects a process within the plurality of processes to monitor the additional process, and the processing unit instructs the selected process to modify the mutually monitoring process policy, wherein the selected process monitors the additional process for termination of execution.

31. A computer program product in a computer-readable medium for mutual process monitoring for detecting a termination of a process within a plurality of processes in a data processing system, comprising:

instructions for establishing a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system;

instructions, responsive to a termination of execution of monitored process, for determining the cause of the execution termination by the monitoring process; and

instructions, responsive to a determination that the monitored process terminated execution in an abnormal manner, for attempting to restart the monitored process by the monitoring process.

32. The computer program product as recited in claim 31, further comprising:

instructions for opening a communications link between the monitoring process and the monitored process; and

instructions for posting a blocking read by the monitoring process in order to detect termination of the monitored process.

33. The computer program product as recited in claim 32, wherein the communications link is a watch FIFO communications link.

34. The computer program product as recited in claim 33, wherein the instructions for posting a blocking read to the watch FIFO communications link is performed in a single thread of execution.

35. The computer program product as recited in claim 32, wherein the monitoring process posts a blocking operation on the communications link that completes the blocking operation on termination of the monitored process.

36. The computer program product as recited in claim 32, wherein the blocking operation is performed in a separate execution thread within the monitoring process in order to allow normal process operation of the monitored process to continue.

37. The computer program product as recited in claim 31, further comprising:

instructions, responsive to a determination that the monitored process terminated execution in a normal manner, for modifying the monitoring policy, wherein the monitoring policy is modified to assign the monitoring process to monitor a different process within the plurality of processes; and

instructions for monitoring of the different process by the monitoring process for termination of execution.

38. A computer program product in a computer-readable medium for inserting an additional process within a plurality of processes containing a mutually monitoring process policy in a data processing system, comprising:

instructions for establishing a monitoring policy, wherein the monitoring policy assigns each process within the plurality of processes to monitor one other process within the plurality of processes, thereby forming a ring of mutually monitoring processes within the data processing system, wherein the data processing system is a single node within a multi-node system;

instructions for receiving a request for an additional process to join the plurality of processes;

instructions, responsive to the additional process joining the plurality of processes, for randomly selecting a process within the plurality of processes to monitor the additional process; and

instructions for modifying the mutually monitoring process policy, wherein the selected process monitors the additional process for termination of execution.

39. The computer program product as recited in claim 38, wherein the additional process monitors the process previously monitored by the selected process.

40. The computer program product as recited in claim 38, wherein the additional process joining the plurality of processes joins as a child of the selected process.

41. The computer program product as recited in claim 38, wherein the request is received by a inter-process communications (IPC) mechanism.

42. The computer program product as recited in claim 38, further comprising:

instructions for locking a state file in a join first-in first-out communications link (FIFO); and

instructions for blocking any other processes attempting to simultaneously join the plurality of processes.

43. The computer program product as recited in claim 42, wherein the communications link is a FIFO communications link.

44. The method as recited in claim 42, wherein blocking any other processes attempting to simultaneously join the plurality of processes is by a file lock.