Thread monitoring using shared memory

• US 7,689,989 B2
• Filed: 12/28/2004
• Issued: 03/30/2010
• Est. Priority Date: 12/28/2004
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP AG (SAP SE)
• Inventors: Werner, Randolf, Dostert, Jan, Fleischer, Christian, Kilian, Frank, Petev, Petio
• Primary Examiner(s): An; Meng-Ai
• Assistant Examiner(s): KESSLER, GREGORY AARON
• Application Number:
  US11/024,394
• Publication Number:
  US 20060143608A1
• Time in Patent Office: 1,918 Days
• Field of Search: 718/1, 718/104, 718/105, 718/100, 709/213, 709/224
• US Class Current: 718/100
• CPC Class Codes: G06F 11/3636 by tracing the execution of the program
• CPC Class Codes: G06F 2209/5018 Thread allocation
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals

Abstract

A system and method to monitor worker threads within a Java virtual machine. At least one program is executed on the Java virtual machine to perform a task. A worker thread of the Java virtual machine is assigned to perform the task. Status information for the worker thread is generated and reported into memory external to the Java virtual machine to enable monitoring of the worker thread from external to the Java virtual machine.

Claims

1. A computer-implemented method, comprising:

operating an application server instance having a plurality of worker nodes, including for each of the plurality of worker nodes,

managing one or more worker threads by a thread manager of the worker node,

executing at least one program on a Java virtual machine (“

JVM”

) of the worker node to service a respective work request received by the application server instance,

assigning to a worker thread of the worker node a task of the respective work request,

the assigned worker thread generating status information based on a performance of the task assigned to the worker thread, and

the assigned worker thread reporting the generated status information into a shared memory of the application server instance, the shared memory external to and shared by the plurality of worker nodes;

wherein a first work request of the work requests received by the application server instance requests a performance of a first task having a stack of multiple subtasks, each subtask of the task corresponding to a respective depth in the stack;

wherein a first worker thread in a first worker node of the plurality of worker nodes is assigned to the first task;

wherein the first worker thread is instantiated by the thread manager of the first worker node to perform both reporting tasks and the subtasks in an interleaved manner;

wherein the performing of reporting tasks includes the first worker thread reporting into the shared memory first status information generated based on a performance of one of the multiple subtasks, the reporting the first status information based on a selectable reporting depth and the corresponding depth in the stack of the one of the multiple subtasks; and

exchanging status information in the shared memory from the application server instance to a monitoring console for monitoring of the application server instance.

2. The method of claim 1, wherein managing one or more worker thread includes creating the first worker thread with a class including at least one reporting method, the reporting method imparting reporting functionality to the first worker thread, the reporting functionality to report the status information into the shared memory.

3. The method of claim 2, further comprising registering a reporting slot within the shared memory to store the status information about the first worker thread.

4. The method of claim 3, further comprising registering a first plurality of reporting slots within the shared memory to store the status information about a corresponding first plurality of worker threads of one of the plurality of worker nodes.

5. The method of claim 4, wherein assigning the first worker thread to the first task comprises assigning the first worker thread from a thread pool of available worker threads, and further comprising:

reporting an idle status into the shared memory for each of the available worker threads idle within the thread pool.

6. The method of claim 1, wherein the first status information includes at least one of a thread state, a thread creation timestamp, a last thread update timestamp, a last task update timestamp, a last subtask update timestamp, a thread pool name, a class name, a thread name, a task name, a subtask name, and a username of a client originating a corresponding one of the work requests.

7. The method of claim 1, wherein the first status information includes a call stack having an execution history of the first worker thread.

8. The method of claim 1, further comprising:

the monitoring console retrieving the first status information from the shared memory; and

transmitting the first status information to the monitoring console to display the first status information.

9. A machine-accessible medium that provides instructions that, if executed by a machine, will cause the machine to perform operations comprising:

operating an application server (“

AS”

) instance having a plurality of worker nodes, including for each of the plurality of worker nodes,

managing one or more worker threads by a thread manager of the worker node,

executing a program on a Java virtual machine (“

JYM”

) of the worker node to service a respective work request received by the AS instance,

assigning to a worker thread of the worker node a task of the respective work request,

the assigned worker thread generating status information based on a performance of the task assigned to the worker thread, and

the assigned worker thread storing the generated status information into a shared memory external to and shared by the plurality of worker nodes;

wherein a first work request of the work requests received by the AS instance requests a performance of a first task having a stack of multiple subtasks, each subtask of the task corresponding to a respective depth in the stack;

wherein the first task is assigned to a first worker thread in a first worker node of the plurality of worker nodes;

wherein the first worker thread is instantiated by the thread manager of the first worker node to perform both reporting tasks and the subtasks in an interleaved manner;

wherein the performing of reporting tasks includes the first worker thread reporting into the shared memory first status information generated based on a performance of one of the multiple subtasks, the reporting the first status information based on a selectable reporting depth and the corresponding depth in the stack of the one of the multiple subtasks; and

exchanging status information in the shared memory from the AS instance to a monitoring console for monitoring of the AS instance.

10. The machine-accessible medium of claim 9, further providing instructions that, if executed by the machine, will cause the machine to perform further operations, comprising:

creating each of the worker threads of the plurality of worker nodes to include a reporting method, wherein the reporting method of each of the worker threads reports corresponding status information into the shared memory.

11. The machine-accessible medium of claim 10, further providing instructions that, if executed by the machine, will cause the machine to perform further operations, comprising:

registering a reporting slot within the shared memory for each of the worker threads, the reporting slot to store the status information corresponding to one of the worker threads.

12. The machine-accessible medium of claim 11, wherein generating the status information includes generating the status information while executing the tasks.

13. A system, comprising:

a server node to execute an application server (“

AS”

) instance having a plurality of worker nodes, the AS instance including logic executable by a processor of the server node to;

for each of the plurality of worker nodes,

manage one or more worker threads by a thread manager of the worker node,

execute a program on a Java virtual machine (“

JVM”

) of the worker node to service a respective work request received by the AS instance;

assign to a worker thread of the worker node a task of the respective work request,

generate with the assigned worker thread status information based on a performance of the task assigned to the worker thread, and

report the generated status information into a shared memory which is external to and shared by the plurality of worker nodes;

wherein a first work request of the work requests received by the AS instance requests a performance of a first task having a stack of multiple subtasks, each subtask of the task corresponding to a respective depth in the stack;

wherein the first task is assigned to a first worker thread in a first worker node of the plurality of worker nodes;

wherein the first worker thread is instantiated by the thread manager of the first worker node to perform both reporting tasks and the subtasks in an interleaved manner;

wherein the performing of reporting tasks includes the first worker thread reporting into the shared memory first status information generated based on a performance of one of the multiple subtasks, the reporting the first status information based on a selectable reporting depth and the corresponding depth in the stack of the one of the multiple subtasks; and

exchange status information in the shared memory from the AS instance to a monitoring console for monitoring of the AS instance.

14. The system of claim 13, wherein the worker threads of the plurality of worker nodes each include reporting functionality to report status information into the shared memory.

15. The system of claim 14, wherein the shared memory includes reporting slots to store the status information, each of the reporting slots to store only status information from one of the worker threads.

16. The system of claim 13, further comprising a client node to execute the monitoring console, the monitoring console including logic executable by a processor of the client node to:

send a status query to the server node;

receive the status information from the server node; and

display the status information in a user readable format.