Computer performance estimation system configured to take expected events into consideration

• US 7,099,799 B2
• Filed: 10/27/2004
• Issued: 08/29/2006
• Est. Priority Date: 10/27/2003
• Status: Expired due to Fees

• Current Assignee: Metricly, Inc. (Virtana Corp.)
• Original Assignee: Netuitive, Inc. (Virtana Corp.)
• Inventors: Huard, Jean-François
• Primary Examiner(s): Raymond, Edward
• Application Number:
  US10/974,062
• Publication Number:
  US 20050091005A1
• Time in Patent Office: 671 Days
• Field of Search: 702/186, 702182-185, 702/188, 700/25, 703/2, 706/21
• US Class Current: 702/186
• CPC Class Codes: G06F 11/3419 by assessing time
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 2201/86 Event-based monitoring

Abstract

The present invention may be embodied as expected event scheduler and processor in an application performance monitoring (APM) services. The expected event scheduler and processor allows the APM system to take scheduled events into account when performing the performance forecasting for the host system. The learned parameters may be based on measured input values including internal measurements, such as data from monitoring agents located within the host computer system, as well as external measurements relating to factors such as computer backup runs, monthly payroll runs, quarterly financial reporting runs, weekends, holidays, weather data, traffic data, advertisements run by the operator of the system or others, promotional events, product releases, news announcements, elections and other natural and demographic factors. The expected events may overlap in time, and the expected event scheduler and processor learns weighing factors for the simultaneously occurring expected events by updating the learned parameters over successive time trials.

Claims

1. In or for an estimation system for a host system operable for receiving input values for successive time trials, computing output values based on the input values and learned parameters, and updating the learned parameters to reflect relationships observed among the input and output values, a method comprising the steps of:

identifying multiple expected events that overlap in time;

provisioning the estimation system with learned parameters for the expected events including weighting factors for the overlapping expected events;

initializing the learned parameters for the expected events; and

running the estimation system to forecast operation of the host system as effected by the expected events while updating the learned parameters including the learned parameters for the expected events.

2. The method of claim 1, wherein the step of running the estimation system comprises the steps of:

(a) receiving measured data corresponding to operation of the host system for a current time trial;

(b) estimating operation of the host system for the current time trial using imputed learned parameters;

(c) predicting operation of the host system for future time trial using forecasted learned parameters;

(d) updating the learned parameters based on the measured data received for the current time trial; and

repeating the steps (a) through (d) for a number of successive time trials.

3. The method of claim 1, wherein the expected events comprise scheduled events.

4. The method of claim 1, wherein the expected events comprise seasonal events.

5. The method of claim 1, wherein the expected events comprise scheduled and seasonal events.

6. The method of claim 1, wherein expected events overlap in time.

7. The method of claim 6, wherein the host system is a computer system and the estimation system forecasts the performance of the computer system.

8. The method of claim 7, wherein the host system comprises a computer system comprises a computer network including application servers, storage servers, routers, and interconnecting data transmission lines.

9. The method of claim 8, wherein the expected events include end-of-month processing events and end-of-quarter processing events.

10. The method of claim 8, wherein the expected events include weekends and holidays.

11. The method of claim 8, wherein the expected events include scheduled backup events.

12. The method of claim 8, wherein the measured data comprises data received from one or more monitoring agents deployed within the computer system.

13. The method of claim 8, wherein the measured data comprises data received from one or more sources external to the computer system.

14. The method of claim 8, wherein the measured data comprises data received from one or more monitoring agents deployed within the computer system and data received from one or more sources external to the computer system.

15. A computer-based estimation system configured to perform the steps recited in claim 1.

16. A computer storage medium storing executable instruction for causing a computer system to implement the method recited in claim 1.

17. In or for an computer performance estimation system operable for receiving input values for successive time trials, computing output values based on the input values and learned parameters, and updating the learned parameters to reflect relationships observed among the input and output values, a method comprising the steps of:

identifying multiple expected events including scheduled events that overlap in time;

provisioning the estimation system with learned parameters for the expected events including weighting factors for the overlapping expected events;

initializing the learned parameters for the expected events; and

running the estimation system to forecast performance of the computer system as effected by the scheduled events while updating the learned parameters including the learned parameters for the scheduled events.

18. The method of claim 17, wherein the step of running the computer performance estimation system comprises the steps of:

(a) receiving measured data from corresponding to operation of the computer system for a current time trial, the measured data including date received from one or more monitoring agents deployed within the computer system;

(b) estimating operation of the computer system for the current time trial using imputed learned parameters;

(c) predicting operation of the computer system for future time trial using forecasted learned parameters;

(d) updating the learned parameters based on the measured data received for the current time trial; and

repeating the steps (a) through (d) for a number of successive time trials.

19. The method of claim 17, wherein:

the expected events further seasonal events including weekends and holidays.

20. The method of claim 19, wherein the host system comprises a computer system comprises a computer network including application servers, storage servers, routers, and interconnecting data transmission lines.

21. The method of claim 20, wherein the expected events further include end-of-month processing events and end-of-quarter processing events.

22. A computer-based estimation system configured to perform the method recited in claim 17.

28. A computer-based estimation system configured to perform the method recited in claim 17.

23. In or for an computer performance estimation system operable for receiving input values for successive time trials, computing output values based on the input values and learned parameters, and updating the learned parameters to reflect relationships observed among the input and output values, a method comprising the steps of:

identifying multiple expected events that overlap in time including scheduled events;

provisioning the estimation system with learned parameters for the expected events including weighting factors for the overlapping expected events;

initializing the learned parameters for the expected events;

running the estimation system to forecast performance of the computer system as effected by the scheduled events while updating the learned parameters including the learned parameters for the scheduled events; and

wherein the step of running the computer performance estimation system comprises the steps of (a) receiving measured data from corresponding to operation of the computer system for a current time trial, the measured data including date received from one or more monitoring agents deployed within the computer system;

(b) estimating operation of the computer system for the current time trial using imputed learned parameters;

(c) predicting operation of the computer system for future time trial using forecasted learned parameters;

(d) updating the learned parameters based on the measured data received for the current time trial; and

repeating the steps (a) through (d) for a number of successive time trials.

24. The method of claim 23, wherein the expected events further seasonal events including weekends and holidays.

25. The method of claim 23, wherein the host system comprises a computer system comprises a computer network including application servers, storage servers, routers, and interconnecting data transmission lines.

26. The method of claim 23, wherein the expected events further include end-of-month processing events and end-of-quarter processing events.

27. A computer storage medium storing executable instruction for causing a computer system to implement the method recited in claim 23.

29. A computer storage medium storing executable instruction for causing a computer system to implement the method recited in claim 23.