System for workload balancing by resetting an average queue depth upon the start of the server instance

• US 8,112,760 B2
• Filed: 03/14/2008
• Issued: 02/07/2012
• Est. Priority Date: 05/28/2003
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Garza, Jose E., Hobson, Stephen J.
• Primary Examiner(s): To, Jennifer
• Assistant Examiner(s): ZHE, MENG YAO
• Application Number:
  US12/048,539
• Publication Number:
  US 20080163249A1
• Time in Patent Office: 1,425 Days
• Field of Search: 718/105
• US Class Current: 718/105
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: G06F 9/546 Message passing systems or structures, e.g. queues

Abstract

The present invention relates to an apparatus and computer program for workload balancing in an asynchronous messaging system. The number of server instances, which process work items from a queue of messages, is controlled based upon that queue'"'"'s average queue depth and one or more predetermined thresholds.

Claims

1. An apparatus, having a processor, for workload balancing in an asynchronous messaging system comprising:

means for obtaining an average queue depth of a message queue;

means for controlling a number of server instances of a server for retrieving messages from the message queue based on the average queue depth and one or more predetermined thresholds;

means for determining if the average queue depth exceeds a first predetermined threshold of the one or more predetermined thresholds;

means for initiating the start of a server instance for retrieving messages from the message queue by placing a trigger message on an initialisation queue in response to determining that the average queue depth exceeds the first predetermined threshold of the one or more predetermined thresholds, wherein the trigger message indicates that the server instance is to be started; and

means for resetting the average queue depth to less than the first predetermined threshold immediately upon the start of the server instance in response to determining that the first predetermined threshold has been exceeded.

2. The apparatus of claim 1, wherein the means for controlling the number of server instances comprises:

means for terminating the server instance when the average queue depth falls below a second predetermined threshold of the one or more predetermined thresholds.

3. The apparatus of claim 2, comprising:

means, responsive to determining that the average queue depth is below the second predetermined threshold, for resetting the average queue depth to greater than the second predetermined threshold immediately upon the termination of the server instance.

4. The apparatus of claim 2, wherein the means for terminating the server instance comprises at least one of:

(i) means for indicating to the server instance that there are no more messages to process on the message queue;

(ii) means for indicating to the server instance that a queue manager, controlling the message queue, is shutting down;

(iii) means for indicating to the server instance that operator intervention is requesting that the server instance shut down;

or

(iv) means for requesting that the server instance shut down.

5. The apparatus of claim 1, comprising:

means for setting a maximum number of server instances that can be active at any one time;

or

means for setting a minimum number of server instances that should be active at any one time.

6. The apparatus of claim 1 wherein the means for obtaining the average queue depth comprises:

means for calculating the average queue depth of the message queue.

7. The apparatus of claim 6 wherein the means for calculating comprises:

means for calculating a time weighted mean average queue depth of the message queue.

8. The apparatus of claim 6, wherein the means for calculating comprises:

means for calculating an exponentially smoothed average queue depth of the message queue.

9. A computer program for workload balancing in an asynchronous messaging system, the computer program stored on a computer recordable medium having computer readable program code, wherein the computer readable program code, when executed in a data processing system, causes the data processing system to:

obtain an average depth of a message queue;

control a number of server instances of a server for retrieving messages from the message queue based on the average queue depth and one or more predetermined thresholds;

determine if the average queue depth exceeds a first predetermined threshold of the one or more predetermined thresholds;

responsive to determining that the average queue depth exceeds the first predetermined threshold, initiate the start of a server instance for retrieving messages from the message queue by placing a trigger message on an initialisation queue, wherein the trigger message indicates that the server instance is to be started; and

responsive to determining that the first predetermined threshold has been exceeded, reset the average queue depth to less than the first predetermined threshold immediately upon the start of the server instance.

10. The computer program of claim 9, wherein the computer readable program code to control the number of server instances further includes computer readable program code that causes the data processing system to:

terminate the server instance when the average queue depth falls below a second predetermined threshold of the one or more predetermined thresholds.

11. The computer program of claim 10, wherein the computer readable program code further causes the data processing system to:

responsive to determining that the average queue depth is below the second predetermined threshold, reset the average queue depth to greater than the second predetermined threshold immediately upon the termination of the server instance.

12. The computer program of claim 10, wherein the computer readable program code to terminate the server instance further includes computer readable program code that causes the data processing system to perform at least one of:

(i) indicating to the server instance that there are no more messages to process on the message queue;

(ii) indicating to the server instance that a queue manager, controlling the message queue, is shutting down;

(iii) indicating to the server instance that operator intervention is requesting that the server instance shut down;

or

(iv) requesting that the server instance shut down.

13. The computer program of claim 9, wherein the computer readable program code further causes the data processing system to perform at least one of:

setting a maximum number of server instances that can be active at any one time;

or

setting a minimum number of server instances that should be active at any one time.

14. The computer program of claim 9, wherein the computer readable program code to obtain the average queue depth further includes computer readable program code that causes the data processing system to:

calculate the average queue depth of the message queue.

15. A system, comprising:

a processor; and

a memory coupled to the processor, wherein the memory comprises instructions which, when executed by the processor, cause the processor to;

obtain an average depth of a message queue;

control a number of server instances of a server for retrieving messages from the message queue based on the average queue depth and one or more predetermined thresholds;

determine if the average queue depth exceeds a first predetermined threshold of the one or more predetermined thresholds;

responsive to determining that the average queue depth exceeds the first predetermined threshold, initiate the start of a server instance for retrieving messages from the message queue by placing a trigger message on an initialisation queue, wherein the trigger message indicates that the server instance is to be started; and

responsive to determining that the first predetermined threshold has been exceeded, reset the average queue depth to less than the first predetermined threshold immediately upon the start of the server instance.

16. The system of claim 15, wherein instructions for controlling the number of server instances further cause the processor to:

terminate the server instance when the average queue depth falls below a second predetermined threshold of the one or more predetermined thresholds.

17. The system of claim 16, wherein the instructions further cause the processor to:

responsive to determining that the average queue depth is below the second predetermined threshold, reset the average queue depth to greater than the second predetermined threshold immediately upon the termination of the server instance.

18. The system of claim 16, wherein the instructions for terminating the server instance further cause the processor to perform at least one of:

(i) indicating to the server instance that there are no more messages to process on the message queue;

(ii) indicating to the server instance that a queue manager, controlling the message queue, is shutting down;

(iii) indicating to the server instance that operator intervention is requesting that the server instance shut down;

or

(iv) requesting that the server instance shut down.

19. The system of claim 15, wherein the instructions further cause the processor to perform at least one of:

set a maximum number of server instances that can be active at any one time;

or

set a minimum number of server instances that should be active at any one time.

20. The system of claim 15, wherein the instructions for obtaining the average queue depth further cause the processor to:

calculate the average queue depth of the message queue.