Providing optimal number of threads to applications performing multi-tasking using threads

• US 7,681,196 B2
• Filed: 11/18/2004
• Issued: 03/16/2010
• Est. Priority Date: 11/18/2004
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Vedula, Venkata Naga Ravikiran
• Primary Examiner(s): An; Meng-Ai
• Assistant Examiner(s): To; Jennifer N
• Application Number:
  US10/904,596
• Publication Number:
  US 20060107261A1
• Time in Patent Office: 1,944 Days
• Field of Search: 709223-226, 709201-203, 712/215, 712/227, 712/235, 712/239, 718100-108, 718102-105, 702/179, 702/187
• US Class Current: 718/104
• 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

In an environment in which applications perform multi-tasking by using threads, the optimum number of threads to be provided to an application is determined based on historical data. Due to the use of the historical data, optimum number of threads can be accurately determined, leading to better response times for service requests, without creating unneeded overhead on system processing the service requests. In one embodiment, various status information (of potentially several prior weeks) such as number of tasks received and pending, system resources status (e.g. idle processor time and aggregate memory usage in the system), is maintained. In addition, a table may be maintained to indicate the optimum number of threads for a given combination of status information values. Thus, based on the expected status scenario, the optimum number of threads can be determined and provided at various points of time.

Claims

1. A method of supporting an application in a system, said method comprising:

maintaining a historical data representing a number of threads required to support execution of said application at different prior time instances;

determining an optimum number of threads to be provided in a near future based on said historical data;

providing said optimum number of threads at a present time to said application in said system;

receiving a plurality of service requests after said providing;

placing a plurality of tasks in a task queue, wherein said plurality of tasks are required to be executed to process said plurality of service requests, wherein each of said threads is designed to examine said task queue and then select and execute any of the pending tasks in said task queue; and

enabling said application to execute said plurality of tasks using said optimum number of threads,

whereby said optimum number of threads are provided to said application before said plurality of service requests are received by said system.

2. The method of claim 1, further comprises storing a mapping which maps of each of a plurality of combination of status information values to a corresponding optimum number of threads, wherein said determining maps a combination of status information values representing a status in the near future to a corresponding optimum number of threads using said mapping.

3. The method of claim 2, wherein said mapping is represented in the form of a table.

4. The method of claim 2, wherein said mapping is represented in the form of a polynomial.

5. The method of claim 2, wherein said historical data is stored in a non-volatile memory.

6. The method of claim 5, wherein said historical data comprises data corresponding to more than one day.

7. The method of claim 2, wherein said status in the near future is determined based on said historical data.

8. The method of claim 1, wherein said providing comprises creating at least one thread, whereby said at least one thread is contained in said optimum number of threads and said application is executed using said at least one thread also,

wherein said present time corresponds to a duration in which said application is being initialized such that said optimum number of threads are provided at the time of initialization of said application.

9. A computer readable medium storing one or more sequences of instructions for causing a system to support an application, wherein execution of said one or more sequences of instructions by one or more processors contained in said system causes said system to perform the actions of:

maintaining a historical data representing a number of threads required to support execution of said application at different prior time instances;

determining an optimum number of threads to be provided in a near future based on said historical data;

providing said optimum number of threads at a present time to said application in said system;

receiving a plurality of service requests after said providing;

placing a plurality of tasks in a task queue, wherein said plurality of tasks are required to be executed to process said plurality of service requests, wherein each of said threads is designed to examine said task queue and then select and execute any of the pending tasks in said task queue; and

enabling said application to execute said plurality of tasks using said optimum number of threads,

whereby said optimum number of threads are provided to said application before said plurality of service requests are received by said system.

10. The computer readable medium of claim 9, further comprises storing a mapping which maps of each of a plurality of combination of status information values to a corresponding optimum number of threads, wherein said determining maps a combination of status information values representing a status in the near future to a corresponding optimum number of threads using said mapping.

11. The computer readable medium of claim 10, wherein said mapping is represented in the form of a table.

12. The computer readable medium of claim 10, wherein said mapping is represented in the form of a polynomial.

13. The computer readable medium of claim 10, wherein said historical data is stored in a non-volatile memory.

14. The computer readable medium of claim 13, wherein said historical data comprises data corresponding to more than one day.

15. The computer readable medium of claim 10, wherein said status in the near future is determined based on said historical data.

16. A digital processing system:

a random access memory (RAM);

a processor;

a computer readable medium to store one or more instructions which are designed to be retrieved into said RAM and to be executed by said processor to perform the actions of;

maintaining a historical data representing a number of threads required to support execution of said application at different prior time instances;

determining an optimum number of threads to be provided in a near future based on said historical data;

providing said optimum number of threads at a present time to said application in said system;

receiving a plurality of service requests after said providing;

placing a plurality of tasks in a task queue, wherein said plurality of tasks are required to be executed to process said plurality of service requests, wherein each of said threads is designed to examine said task queue and then select and execute any of the pending tasks in said task queue; and

enabling said application to execute said plurality of tasks using said optimum number of threads,

whereby said optimum number of threads are provided to said application before said plurality of service requests are received by said digital processing system.