Mobile agents for fault diagnosis and correction in a distributed computer environment

• US 6,460,070 B1
• Filed: 06/03/1998
• Issued: 10/01/2002
• Est. Priority Date: 06/03/1998
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Turek, John J. E., Vetter, Brian Jay
• Primary Examiner(s): Rinehart, Mark H.
• Assistant Examiner(s): Thompson, Marc D.
• Application Number:
  US09/089,962
• Time in Patent Office: 1,581 Days
• Field of Search: 709/202, 709/223, 709/224, 709/239, 709/240, 709/206, 709/203, 370/241, 370/242, 714/4, 714/25, 714/46
• US Class Current: 709/202
• CPC Class Codes: G06F 11/0709 in a distributed system consisting of a plurality of standalone computer nodes, e.g. clusters, client-server systems
• CPC Class Codes: G06F 11/079 Root cause analysis, i.e. error or fault diagnosis in a hardware test environment G06F11/22; in a software test environment G06F11/36
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: H04L 41/046 comprising network management agents or mobile agents therefor
• CPC Class Codes: H04L 41/0677 Localisation of faults
• CPC Class Codes: H04Q 3/0075 Fault management techniques

Abstract

A large distributed enterprise includes computing resources that are organized into one or more managed regions, each region being managed by a management server servicing one or more gateway machines, with each gateway machine servicing a plurality of endpoint machines. A method of diagnosing a fault in such an environment begins by deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment at each of the endpoint machines. In response to occurrence of the fault, a software agent is selected, the software agent being executable by the runtime environment at an endpoint machine. The selected software agent is then deployed into the computer network to diagnosis the fault. If the location of the fault is indeterminate, the software agent migrates to the location by gathering information about the fault as it traverses the network.

Claims

1. A method of diagnosing a fault in a large, distributed computer network having a management server servicing a set of gateway machines, each of the gateway machines servicing a set of endpoint machines, the method comprising the steps of:

deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment at each of the endpoint machines;

in response to occurrence of the fault, selecting a software agent, the software agent being executable by the runtime environment at an endpoint machine; and

deploying the selected software agent into the computer network to diagnose the fault.

2. The method as described in claim 1 further including the step of having the software agent attempt to correct the fault.

3. The method as described in claim 2 wherein the step of having the software agent attempt to correct the fault includes utilizing code of the software agent to fix the fault.

4. The method as described in claim 2 wherein the step of having the software agent attempt to correct the fault includes requesting additional code to be dispatched from another node in the network.

5. The method as described in claim 4 wherein the additional code is another software agent.

6. The method as described in claim 1 wherein the selecting step includes identifying a given characteristic of the fault and selecting the software agent based on the given characteristic.

7. The method as described in claim 6 wherein the step of selecting the software agent further includes assembling a set of tasks necessary to diagnosis the fault.

8. The method as described in claim 1 wherein the runtime environment comprises a runtime engine and the software agent is a set of one or more tasks executable by the runtime engine.

9. The method as described in claim 8 wherein the computer network includes the Internet, the runtime engine is associated with a Web browser and the software agent is an applet.

11. A method of diagnosing a fault in a large, distributed computer network having a management server servicing a set of gateway machines, each of the gateway machines servicing a set of endpoint machines, the method comprising the steps of:

deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment at each of the endpoint machines;

in response to occurrence of the fault, selecting a software agent, the software agent being executable by the runtime environment at an endpoint machine; and

deploying the selected software agent into the computer network to diagnosis the fault;

wherein the step of deploying the selected software agent comprises the steps of;

migrating the selected software agent across a given set of nodes in the computer network;

at each node at which the selected software agent is received;

(a) determining whether the fault originated from the node;

(b) if the fault did not originate from the node, identifying a given subset of nodes associated with the node that remain candidates for locating the fault;

(c) deploying the software agent to the given subset; and

(d) repeating steps (a)-(c) until a location of the fault is determined.

12. A method of diagnosing and correcting a fault in a large, distributed computer network in which a management infrastructure is supported, comprising the steps of:

in response to occurrence of the fault, selecting at least one software agent from a set of software agents useful in diagnosing network events;

migrating the selected software agent across a given set of nodes in the computer network until the software agent arrives at the location of the fault; and

diagnosing and correcting the fault.

10. The method as described in claim 12 wherein the fault is diagnosed and corrected using code in the software agent.

13. The method as described in claim 12 wherein the management infrastructure includes a runtime environment deployed at each node of the computer network and the software agent is a set of one or more tasks executed by the runtime environment.

14. The method as described in claim 13 wherein the computer network includes the Internet, the runtime environment is associated with a Web browser and the software agent is an applet.

15. The method as described in claim 12 wherein the fault is corrected by dispatching additional code to the location.

16. The method as described in claim 15 wherein the additional code is another software agent.

17. A method of diagnosing and correcting a fault in a large, distributed computer network in which a management infrastructure is supported, comprising the steps of:

in response to occurrence of the fault, selecting at least one software agent from a set of software agents useful in diagnosing network events;

migrating the selected software agent across a given set of nodes in the computer network until the software agent arrives at the location of the fault; and

diagnosing and correcting the fault;

wherein the selecting step includes identifying a given characteristic of the fault and selecting the software agent based on the given characteristic.

18. The method as described in claim 17 wherein the selecting step includes assembling a set of tasks necessary to diagnose the fault.

19. In a large distributed enterprise having a management server servicing a set of gateway machines, each of which services a set of endpoint machines, a diagnostic method comprising the steps of:

deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment at given endpoint machines;

in response to occurrence of a given event, selecting a software agent, the software agent being executable by the runtime environment at an endpoint machine;

deploying the selected software agent into the computer network to diagnosis the event.

20. The diagnostic method as described in claim 19 wherein the event is a fault.

21. In a large distributed enterprise having a management server servicing a set of gateway machines, each of which services a set of endpoint machines, a diagnostic method comprising the steps of:

deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment at given endpoint machines;

in response to occurrence of a given event selecting a software agent, the software agent being executable by the runtime environment at an endpoint machine;

deploying the selected software agent into the computer network to diagnosis the event;

wherein the software agent is deployed by;

migrating the selected software agent across a given set of nodes in the computer network; and

at each node at which the selected software agent is received;

(a) determining whether the fault originated from the node;

(b) if the fault did not originate from the node, identifying a given subset of nodes associated with the node that remain candidates for locating the fault;

(c) deploying the software agent to the given subset; and

(d) repeating steps (a)-(c) until a location of the fault is identified.

22. Apparatus connectable into a large distributed enterprise having a management server servicing a set of gateway machines, each of which services a set of endpoint machines, comprising:

means for deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment for use at each of the endpoint machines;

means, responsive to occurrence of a given event, for selecting a software agent to diagnose the event, the software agent being executable by the runtime environment at an endpoint machine; and

means for deploying the selected software agent into the computer network to diagnosis the event.

23. The apparatus as described in claim 22 wherein the given event is a fault and the software agent is deployed to diagnose and correct the fault.

24. Apparatus connectable into a large distributed enterprise having a management server servicing a set of gateway machines, each of which services a set of endpoint machines, comprising:

means for deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment for use at each of the endpoint machines;

means, responsive to occurrence of a given event, for selecting a software agent to diagnose the event, the software agent being executable by the runtime environment at an endpoint machine; and

means for deploying the computer network to diagnosis the event;

wherein the selecting means includes;

means for parsing information about the event to identify one or more characteristics of the event; and

means responsive to the parsing means for correlating the one or more event characteristics with information about prior events to identify the software agent.

25. A computer program product in computer-readable media for use in a computer having a processor, a memory, and means for connecting the computer into a large distributed enterprise having a management server servicing a set of gateway machines, the computer connectable to one of the gateway machines, the computer program product comprising:

a runtime environment downloaded to the computer during a deployment operation; and

a software agent deployed to the computer during a diagnostic operation and being executable by the runtime environment to diagnose a given condition associated with the computer.

26. The computer program product as described in claim 25 wherein the software agent includes one or more tasks selected in response to the given condition.

27. A computer program product in computer-readable media for use in a computer having a processor, a memory, and means for connecting the computer into a large distributed enterprise having a management server servicing a set of gateway machines, the computer connectable to one of the gateway machines, the computer program product comprising:

a runtime environment downloaded to the computer during a deployment operation; and

a software agent deployed to the computer during a diagnostic operation and being executable by the runtime environment to diagnose a given condition associated with the computer;

wherein the runtime environment is associated with a browser of the computer and the software agent is a Java applet.

28. A computer program product in computer-readable media for use in a computer having a processor, a memory, and means for connecting the computer into a large distributed computer network, the computer network having a management server servicing a set of gateway machines, each of which services a set of endpoint machines, the computer program product comprising:

means for deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment for use at each of the endpoint machines;

a set of software agents, each of the software agents comprising a set of one or more tasks;

means, responsive to occurrence of a given event, for selecting a software agent to diagnose the event, the software agent being executable by the runtime environment at an endpoint machine; and

means for deploying the selected software agent into the computer network to diagnosis the event.

29. The computer program product as described in claim 28 wherein the given event is a fault and the software agent is deployed to diagnose and correct the fault.

30. A computer program product in computer-readable media for use in a computer having a processor, a memory, and means for connecting the computer into a large distributed computer network, the computer network having a management server servicing a set of gateway machines, each of which services a set of endpoint machines, the computer program product comprising:

means for deploying a management infrastructure throughout the computer network, the management infrastructure including a runtime environment for use at each of the endpoint machines;

a set of software agents, each of the software agents comprising a set of one or more tasks;

means, responsive to occurrence of a given event, for selecting a software agent to diagnose the event, the software agent being executable by the runtime environment at an endpoint machine, and

means for deploying the selected software agent into the computer network to diagnosis the event;

wherein the selecting means includes;

means for parsing information about the event to identify one or more characteristics of the event; and

means responsive to the parsing means for correlating the one or more event characteristics with information about prior events to identify the software agent.