Method and device for forecasting computational needs of an application

• US 20090240476A1
• Filed: 03/24/2008
• Published: 09/24/2009
• Est. Priority Date: 03/24/2008
• Status: Active Grant

• Current Assignee: ServiceNow Incorporated
• Original Assignee: AT&T Services Incorporated (AT&T, Inc.)
• Inventors: Joshi, Ajay, Estela, Angelo, Azam, Muhammad, Lu, David, Kshirsagar, Madhukar
• Granted Patent: US 8,180,716 B2
• Time in Patent Office: Days
• US Class Current: 703/2
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3442 for planning or managing the needed capacity
• CPC Class Codes: G06F 11/3452 Performance evaluation by statistical analysis
• CPC Class Codes: G06F 11/3476 Data logging G06F11/14, G06F11/2205 take precedence
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system

Abstract

The invention comprises methods and devices of forecasting future computational needs of an application based on input load, where the application comprises a plurality of processes executed on a plurality of computing devices. The method of the invention proceeds by monitoring at least a computational load characteristic of at least a first process executed on a first computing device and a second process executed on a second computing device. A mathematical relationship between input load and the computational load characteristic is established, and future computational needs are forecasted based on the established mathematical relationship.

Claims

1. A method for predicting future computational need for an application, said application comprising a plurality of processes wherein at least one first process of said application is executed on a first computing device and at least one second process of said application is executed on a second computing device;

said method comprising the steps of;

monitoring a computational load characteristic of the first process and of the second process;

establishing a mathematical relationship between input load and said computational load characteristic for each of said first process and said second process; and

forecasting a future computational need for at least one of said first process and said second process, based on said established mathematical relationship.

2. The method of claim 1, wherein each process of said plurality of processes is executed on a separate computing device.

3. The method of claim 1, wherein a plurality of said processes is executed serially.

4. The method of claim 1, wherein at least one computing device executes a process for multiple applications.

5. The method of claim 4, wherein an aggregate value of computational load is measured on a said at least one computing device based on computational load of each application of said multiple applications.

6. The method of claim 1, wherein said monitoring comprises monitoring of average computational load.

7. The method of claim 1, wherein said monitoring comprises monitoring of peak computational load.

8. The method of claim 1, wherein said mathematical relationship is a linear relationship.

9. The method of claim 1, wherein said mathematical relationship is a logarithmic relationship.

10. The method of claim 1 further comprising a step of alerting a user when a monitored computational load characteristic exceeds a predefined threshold.

11. The method of claim 1, wherein said monitored computational load characteristic is selected from the group consisting of cpusecs, memory usage, network usage, and disk usage.

12. A device for predicting future computational need for an application executed across a plurality of computing devices, said device comprising:

an application poller configured to receive computational load characteristic data from a plurality of processes of said application, wherein at least a first process of said plurality is executed on a first computing device and at least a second process of said plurality is executed on a second computing device;

an aggregator configured to conduct mathematical calculations on said computational load characteristic data;

a correlating engine configured to establish a mathematical relationship between input load and results of said mathematical calculations; and

an enterprise computing engine configured to forecast future computational need for at least one of said first and second processes based on said established mathematical relationship.

13. The device of claim 12, wherein each process of said plurality of processes is executed on a separate computing device.

14. The device of claim 12, wherein said plurality of processes is executed serially.

15. The device of claim 12, wherein at least one computing device executes a process for multiple applications.

16. The method of claim 15, wherein an aggregate value of computational load is measured on a said at least one computing device based on computational load of each application of said multiple applications.

17. The device of claim 12, wherein said mathematical calculations comprise a computation of average computational load.

18. The device of claim 12, wherein said mathematical calculations comprise a computation of peak computational load.

19. The device of claim 12, wherein said mathematical relationship is a linear relationship.

20. The device of claim 12, wherein said mathematical relationship is a logarithmic relationship.

21. The device of claim 12, further comprising means for alerting a user when a monitored computational load characteristic exceeds a predefined threshold.

22. The device of claim 12, wherein said computational load characteristic is selected from the group consisting of cpusecs, memory usage, network usage, and disk usage.