Application processing allocation in a computing system

• US 10,241,843 B2
• Filed: 11/03/2017
• Issued: 03/26/2019
• Est. Priority Date: 12/01/2009
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Kaplinger, Todd E., Shook, Aaron K.
• Primary Examiner(s): Mejia, Anthony
• Application Number:
  US15/803,531
• Publication Number:
  US 20180060139A1
• Time in Patent Office: 508 Days
• Field of Search: 709226
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system

Abstract

A method for allocating processing of an application performed by a computing system made up of a plurality of interconnected physical computing devices includes executing an application on a first application server associated with the computing system, the application having a number of modular software components; while executing the application, measuring processing resources consumed by one of the modular software components; and in response to one of the modular software components consuming an amount of processing resources defined by a criterion, deploying the one of the modular software components to a second application server associated with the computing system.

Claims

1. A computer program product for allocating processing of an application performed by a computing system comprising a plurality of physical computing devices, the computer program product comprising:

a non-transitory computer readable storage medium having computer readable code embodied therewith, the computer readable program code, when executed, causing a first application server with a network interface to provide communication with a second application server to;

execute an application, the application comprising a plurality of modular software components, with the plurality of modular software components including a first modular software component and a second modular software component;

during an execution period of the application, measure a first set of performance metric value(s) for the first modular software component with respect to the execution period;

measure a second set of performance metric value(s) for the second modular software component with respect to the execution period;

determine that the first set of performance metric value(s) indicates less favorable performance than the second set of performance metric value(s); and

responsive to the determination that the first set of performance metric values(s) indicates less favorable performance, transfer the first modular software component to the second application server to be processed by a processor associated with the second application server;

wherein transferring the first modular software component to be processed by the second application server is conducted by a remote request dispatcher (RRD).

2. The computer program product of claim 1, wherein a measured performance metric is latency.

3. The computer program product of claim 1, wherein transferring the first modular software component to the second application server comprises activating a context sharing mechanism.

4. The computer program product of claim 1, wherein the second application server is one of a plurality of application servers, the first modular software component being transferred to the second application server responsive to the second application server consuming a smaller amount of its own processing resources than any other application server in the plurality of application servers.

5. The computer program product of claim 1, wherein a virtual machine associated with at least one of the first application server and the second application server is used to process the application.

6. The computer program product of claim 1, wherein the code for transferring the first modular software component to the second application server comprises code to:

identify a remote resource on a remote module on the second application server based on a reference associated with a resource on the first application server;

create an RRD request object;

send the RRD request object to the second application server;

receive a response to the RRD request from the second application server; and

present contents of the response including a description of the available contents, response output, and response state of the second application server to be made available to the resource of the first application server.

7. A computer program product for allocating processing of an application performed by a computing system comprising a plurality of physical computing devices, the computer program product comprising:

a non-transitory computer readable storage medium having computer readable code embodied therewith, the computer readable program code, when executed, causing a first application server with a network interface to provide communication with a second application server to;

execute an application, the application comprising a plurality of modular software components;

during an execution period of the application, measure a performance metric value for each of a first modular software component and second modular software component of the application;

determine when the measured performance metric values indicate a less favorable performance for the first modular software component than the second modular software component; and

responsive to the determination that the measured performance metric values indicate a less favorable performance for the first modular software component than the second modular software component, re-deploy the first modular software component to the second application server to be processed by a processor associated with the second application server;

wherein re-deploying the first modular software component to be processed by the second application server is conducted by a remote request dispatcher (RRD).

8. The computer program product of claim 7, wherein a measured performance metric is latency.

9. The computer program product of claim 7, wherein re-deploying the first modular software component to the second application server comprises activating a context sharing mechanism.

10. The computer program product of claim 7, wherein the second application server is one of a plurality of application servers, the first modular software component being re-deployed to the second application server responsive to the second application server consuming a smaller amount of its own processing resources than any other application server in the plurality of application servers.

11. The computer program product of claim 7, wherein a virtual machine associated with at least one of the first application server and the second application server is used to process the application.

12. The computer program product of claim 7, wherein the code for re-deploying the first modular software component to the second application server comprises code to:

identify a remote resource on a remote module on the second application server based on a reference associated with a resource on the first application server;

create an RRD request object;

send the RRD request object to the second application server;

receive a response to the RRD request from the second application server; and

present contents of the response including a description of the available contents, response output, and response state of the second application server to be made available to the resource of the first application server.

13. A method for allocating processing of an application performed by a computing system comprising a plurality of interconnected physical computing devices, the method comprising:

executing an application, the application comprising a plurality of modular software components;

during an execution period of the application, measuring a performance metric value for each of a first modular software component and second modular software component of the application;

determining when the measured performance metric values indicate a less favorable performance for the first modular software component than the second modular software component; and

responsive to the determination that the measured performance metric values indicate a less favorable performance for the first modular software component than the second modular software component, re-deploying the first modular software component to the second application server to be processed by a processor associated with the second application server;

wherein re-deploying the first modular software component to be processed by the second application server is conducted by a remote request dispatcher (RRD).

14. The method of claim 13, wherein a measured performance metric is latency.

15. The method of claim 13, wherein re-deploying the first modular software component to the second application server comprises activating a context sharing mechanism.

16. The method of claim 13, wherein the second application server is one of a plurality of application servers, the first modular software component being re-deployed to the second application server responsive to the second application server consuming a smaller amount of its own processing resources than any other application server in the plurality of application servers.

17. The method of claim 13, wherein a virtual machine associated with at least one of the first application server and the second application server is used to process the application.

18. The method of claim 13, wherein re-deploying the first modular software component to the second application server further comprises:

identifying a remote resource on a remote module on the second application server based on a reference associated with a resource on the first application server;

creating an RRD request object; and

sending the RRD request object to the second application server.

19. The method of claim 13, further comprising:

receiving a response to the RRD request from the second application server; and

presenting contents of the response including a description of the available contents, response output, and response state of the second application server to be made available to the resource of the first application server.

20. The method of claim 13, further comprising using a tolerance for expected latency in determining whether to deploy the first application server to the second application server.