Process monitoring in a computer system

• US 6,640,203 B2
• Filed: 04/05/2001
• Issued: 10/28/2003
• Est. Priority Date: 10/09/1998
• Status: Expired due to Term

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Brown, Roger S., Roles, Karen C., Applebaum, Simon G.
• Primary Examiner(s): Barlow, John
• Assistant Examiner(s): DOUGHERTY, ANTHONY T
• Application Number:
  US09/828,495
• Publication Number:
  US 20020116151A1
• Time in Patent Office: 936 Days
• Field of Search: 702/186
• US Class Current: 702/186
• CPC Class Codes: G06F 11/0715 in a system implementing multitasking multitasking per se G06F9/46
• CPC Class Codes: G06F 11/1438 Restarting or rejuvenating
• CPC Class Codes: G06F 11/16 Error detection or correction of the data by redundancy in hardware
• CPC Class Codes: G06F 11/1641 where the comparison is not performed by the redundant processing components
• CPC Class Codes: G06F 11/28 by checking the correct order of processing G06F11/08 - G06F11/26 take precedence; monitoring patterns of pulse trains H03K5/19
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2201/845 Systems in which the redundancy can be transformed in increased performance
• CPC Class Codes: G06F 9/4843 by program, e.g. task dispatcher, supervisor, operating system

Abstract

The health of a process is monitored in a computer system by a process monitor. The monitored process (a configuration management system daemon (CMSD)) is not a child of the process monitor. The process monitor uniquely determines the identity of a monitored process and verifies the correct operation of the monitored process. In the absence of verification of the correct operation of the monitored process, the monitored process is caused to initiate. On successful initiation of the monitored process, the monitored process is uniquely identified to the system and is detached from the process monitor. Each monitored process is arranged to write, on initiation, its unique process identification information (PID) to a file, which file is then accessed by the process monitor to identify the process monitor. The process monitor can interrogate the operating system to verify correct operation of the CMSD. As an alternative, the process monitor could test whether the CMSD is functioning by making service requests to the CMSD.

Claims

1. A method of monitoring a process in a computer system using a process monitor, where the monitored process is not a child of the process monitor;

the method including steps of;

a) uniquely determining the identity of a monitored process and verifying correct operation of the monitored process;

b) in the absence of verification of the correct operation of the monitored process, causing the monitored process to restart, the restarted process initially checking that it is operable and, if it is operable, indicating to the process monitor that it is operable and backgrounding itself, the monitored process, on successful restart, uniquely identifying itself to the computer system.

2. The method of claim 1, wherein step (b) further comprises the restarted monitored process writing process identification information uniquely identifying itself to a predetermined location.

3. The method of claim 1, wherein step (b) further comprises:

in the absence of an operable indication from the restarted monitored process, the process monitor issuing an error message and preventing further attempts to restart the monitored process.

4. The method of claim 1, wherein the process monitor is a monitor process.

5. The method of claim 1, wherein the monitored process is a daemon process.

6. The method of claim 1, wherein the step (a) comprises attempting access to a predetermined location for process identification information uniquely identifying a monitored process.

7. The method of claim 6, wherein the predetermined location is a predetermined file.

8. The method of claim 1, wherein the monitored process is a configuration management system daemon.

9. The method of claim 8, wherein the configuration management system daemon responds to object definitions.

10. The method of claim 1, wherein step (a) comprises using the unique identity of the monitored process to verify correct operation of the monitored process.

11. The method of claim 10, wherein step (a) comprises interrogating an operating system of the computer system to verify correct operation of the monitored process.

12. The method of claim 10, wherein step (a) comprises requesting a service from the monitored process for verifying correct operation thereof.

13. A method of restarting a process to be monitored in a computer system, the method comprising:

spawning a new process; and

the new process checking that it is operable and, in response to determining that the new process is operable;

uniquely identifying itself to the computer system; and

causing an existing monitored process to terminate, whereby the new process becomes the monitored process.

14. The method of claim 13, comprising recording unique process identification information for the new process at a predetermined location for uniquely identifying the new process.

15. The method of claim 13, wherein the existing monitored process is a first invocation of a configuration management system daemon operable on a first set of object definitions, and the new process is a second invocation of a configuration management system daemon operable on a second set of object definitions.

16. The method of claim 13, wherein the restarted monitored process is selected from:

a new process, an upgrade version of a process, or an existing process.

17. A computer system comprising a process to be monitored, the process to be monitored being configured, on successful initiation, uniquely to identify itself to the system, and a process monitor configured:

uniquely to determine the identity of a monitored process;

to verify correct operation of the monitored process; and

, in the event of being unable to verify correct operation of the monitored process, to cause the monitored process to restart, wherein the monitored process, on successful restart, initially checks that it is operable and, if it is operable, indicates to the process monitor that it is operable, backgrounds itself and, uniquely identifies itself to the system.

18. The computer system of claim 17, wherein the process monitor is a monitor process.

19. The computer system of claim 17, wherein the monitored process is a daemon process.

20. The computer system of claim 17, comprising at least one processor set for executing the process monitor and the monitored process, and memory forming said storage.

21. The computer system of claim 17, wherein the restarted monitored process is selected from a new process, an upgrade version of a process, or an existing process.

22. The computer system of claim 17, wherein the process monitor is configured to attempt access a predetermined storage location of the computer system for process identification information uniquely identifying a monitored process.

23. The computer system of claim 22, wherein the predetermined location is a predetermined file held in computer memory.

24. The computer system of claim 17, wherein the restarted process to be monitored writes processor identification information uniquely identifying itself to a predetermined location after verifying that it is able to function successfully.

25. The computer system of claim 24, wherein the process monitor is configured to issue an error message and to prevent further attempts to restart the monitored process in the absence of an indication from the restarted monitored process that it is operable.

26. The computer system of claim 17, wherein the process monitor is configured to use the unique identity of the monitored process to verify correct operation of the monitored process.

27. The computer system of claim 26, wherein the process monitor is configured to interrogate an operating system of the computer system to verify correct operation of the monitored process.

28. The computer system of claim 26, wherein the process monitor is configured to request a service from the monitored process for determining correct operation thereof.

29. The computer system of claim 17, wherein the monitored process is a configuration management system daemon.

30. The computer system of claim 29, wherein the configuration management system daemon responds to object definitions.

31. The computer system of claim 30, wherein the computer system is a fault-tolerant computer system including a plurality of processing sets operable in lockstep, the monitored process being a configuration management system daemon responsive to configuration system definitions representative of elements of the fault tolerant computer system.

32. A computer program on a carrier medium for the computer system of claim 17, the computer program forming a configuration management system comprising program code configured, on being initiated by a process monitor, to check that it is operable;

and, if so, to provide an indication of this to the process monitor prior to detaching itself from the process monitor, so as not to be a child of the process monitor.

33. The computer program of claim 32, comprising a configuration management system daemon.

34. The computer program of claim 32, comprising program code forming a configuration system management daemon responsive to configuration management system definitions representative of computer system resources.

35. The computer program of claim 32, further configured to identify itself to the system storage after verifying that it is operable.

36. The computer program of claim 35, further configured to record process identification information uniquely identifying itself at a predetermined location in storage for uniquely identifying itself to the system.

37. A computer program on a carrier medium, the computer program forming a process monitor including program code configured uniquely to determine the identity of a monitored process, to verify correct operation of the monitored process, and, in the event of being unable to verify correct operation of the monitored process, to cause the monitored process to restart, such that the monitored process is not a child of the process monitor, and wherein the monitored process, on successful restart, initially checks that it is operable and, if it is operable, indicates to the process monitor that it is operable, backgrounds itself, and uniquely identifies itself to the system.

38. The computer program of claim 37, wherein the process monitor is configured to attempt access to a predetermined storage location of the computer system for process identification information uniquely identifying a monitored process.

39. The computer program of claim 37, wherein the process monitor is configured to issue an error message and to prevent further attempts to restart the monitored process in the absence of an indication from the restarted monitored process that it is operable.

40. The computer program of claim 37, wherein the restarted monitored process is selected from:

a new process, an upgrade version of a process, or an existing process.

41. The computer program of claim 37, wherein the process monitor is a monitor program carried by a carrier medium.

42. The computer program of claim 37, wherein the process monitor is configured to use the unique identity of the monitored process to verify correct operation of the monitored process.

43. The computer program of claim 42, wherein the process monitor is configured to interrogate an operating system of the computer system to verify correct operation of the monitored process.

44. The computer program of claim 42, wherein the process monitor is configured to request a service from the monitored process for determining correct operation thereof.