Method and System for Achieving Better Efficiency in a Client Grid Using Node Resource Usage and Tracking

• US 20090276519A1
• Filed: 05/05/2008
• Published: 11/05/2009
• Est. Priority Date: 05/05/2008
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Sweeney, William, Alimi, Richard A., Zhou, Nianjun
• Granted Patent: US 8,903,981 B2
• Time in Patent Office: Days
• US Class Current: 709/224
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 9/5072 Grid computing

Abstract

A system, method and computer program product for distributing task assignments on a computer network comprising a client grid having at least one server coupled to at least one client node and a plurality of client computers coupled to the client node through a plurality of monitoring agents. Each monitoring agent collects data regarding the resources a particular client computer makes available to the grid and transmits the data to the grid server when the client computer requests a grid task. The system generates a resource probability distribution based on the historical computing resource data and employs a scheduling algorithm to distribute grid tasks to the client computers using at least the probability distribution.

Claims

1. A system for distributing task assignments on a computer network comprising:

a client grid having at least one server coupled to at least one client node,

a plurality of client computers coupled to said client node through a plurality of monitoring agents;

wherein each said monitoring agent collects data regarding the available resources a particular client computer and transmits said data to the grid server when said client machine request grid task.

2. The system of claim 1 wherein said monitoring agent resides on said client computer.

3. The system of claim 2 wherein said monitoring agent collects data relating to the host client computer'"'"'s computing resources available to the grid.

4. The system of claim 3 wherein said host client computer'"'"'s computing resources available to the grid includes the client computer'"'"'s CPU availability.

5. The system of claim 3 wherein said host client computer'"'"'s computing resources available to the grid includes the client computer'"'"'s memory availability.

6. The system of claim 3 wherein said host client computer'"'"'s computing resources available to the grid includes the amount of available network bandwidth.

7. The system of claim 3 wherein said host client computer'"'"'s computing resources available to the grid includes the amount of time dedicated to grid tasks.

8. The system of claim 3 wherein said resource availability data includes computer resource usage data.

19. The system of claim 1 wherein said monitoring agent transmits said client computer resource availability data at predetermined intervals.

20. The system of claim 1 wherein said system server receives said transmitted historical computing resource availability data for a plurality of client computers coupled to said grid, generates a resource probability distribution based on the historical computing resource data, and employs a scheduling algorithm to distribute grid tasks to said plurality of client computers.

9. A method of efficiently distributing task assignments from a server to client computers on a computer grid comprising:

monitoring the resource availability of a plurality of client computers coupled to a grid;

transmitting resource availability data from said plurality of client computers to a grid server;

collecting historical computing resource availability data for said plurality of client computers coupled to said grid;

generating a resource probability distribution based on the historical computing resource data;

employing a scheduling algorithm to distribute grid tasks to said client computers using as least said probability distribution.

10. The method of claim 9 further comprising at least one monitoring agent, said at least monitoring agent residing on at least one client computer coupled to the grid and monitoring the computing resources that said at least one client computer makes availability to said grid.

11. The method of claim 10 wherein at least one monitoring agent resides on each client computer coupled to the said grid, monitoring the computing resources that each of said at least one client computer makes availability to said grid.

12. The method of claim 9 wherein said resource availability data includes said host client computer'"'"'s CPU resources available to the grid.

13. The method of claim 9 wherein said resource availability data includes said host client computer'"'"'s memory resources available to the grid.

14. The method of claim 9 wherein said resource availability data includes said host client computer'"'"'s network bandwidth resources available to the grid.

15. The method of claim 9 wherein said resource availability data includes the amount of time said host client computer dedicates to grid tasks.

16. A computer program product comprising computer-usable medium having a computer-usable program code for efficiently distributing task assignments from a server to client computers on a computer grid said computer program product comprising:

computer-usable program code for monitoring the resource availability of a plurality of client computers coupled to a grid;

computer-usable program code for transmitting resource availability data from said plurality of client computers to a grid server;

computer-usable program code for collecting historical computing resource availability data for a plurality of client computers coupled to said grid;

computer-usable program code for generating a resource probability distribution based on the historical computing resource data;

computer-usable program code employing a scheduling algorithm to distribute grid tasks to said client computers using as least said probability distribution.

17. The computer program product of claim 16 wherein said computer-usable program code for monitoring the resource availability of a plurality of client computers coupled to a grid further comprises at least one monitoring agent, said at least monitoring agent residing on at least one client computer coupled to the grid and monitoring the computing resources that said at least one client computer makes availability to said grid.

18. The computer program product of claim 17 wherein at least one monitoring agent resides on each client computer coupled to the said grid, monitoring the computing resources that each of said at least one client computer makes availability to said grid.