System, method and program product for forecasting the demand on computer resources

• US 20060010101A1
• Filed: 07/05/2005
• Published: 01/12/2006
• Est. Priority Date: 07/08/2004
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Suzuki, Yasuhiro, Yamamoto, Katsuyoshi, Yamamoto, Naoya
• Granted Patent: US 7,415,453 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/3409 for performance assessment
• 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 2201/87 Monitoring of transactions
• CPC Class Codes: G06F 2201/875 Monitoring of systems including the internet

Abstract

An automated system and method for predicting a future resource utilization tendency by analyzing a change in resource usage on the basis of a real operation condition in a computer system. A CPU utilization estimating unit acquires a resource log and a transaction log, and estimates utilization for each transaction, and information on the number of executions for each transaction, a processing amount predicting unit performs a statistical time-series analysis on the basis of time-series data of the number of executions for each transaction included in the transaction log and predicts an amount of processing for each type of a transaction, and a resource utilization predicting unit predicts the computer resource utilization or demand on from the estimated resource utilization for each transaction and the future predicted amount of processing for each transaction.

Claims

1. A system for predicting a demand on a computer resource, comprising:

first means for acquiring a first log on a use of the resource and a second log on transactions using the resource;

second means for estimating resource utilization for each transaction on the basis of information on resource use included in the first log and information on the number each transaction processed included in the second log;

third means for acquiring the second log and, performing a statistical time-series analysis on the basis of time-series data of the number of each transaction processed, and predicting a future amount of processing volume for each transaction type; and

fourth means for predicting a computer resource demand on the basis of the resource utilization for each transaction estimated by said second means and the future amount of processing for each transaction predicted by said third means.

2. The system according to claim 1, wherein said second means obtains resource use at a plurality of times and the number each transaction at the corresponding time from the first log and the second log and estimates the resource utilization for each transaction by a multiple linear regression analysis.

3. The system according to claim 1, wherein said third means identifies a time-series model by a statistical time-series analysis and predicts the amount of processing for each transaction using the time-series model.

4. The system according to claim 3, wherein said third means excludes a part of the time-series data, identifies the time-series model on the basis of remaining data, predicts the number of processes corresponding to the excluded part of the time-series data from the identified time-series model, and compares a result of the prediction with actual data of the excluded part to verify the time-series model.

5. The system according to claim 4, wherein said third means excludes the last data of a certain period in the time-series data and uses the data of the certain period for verifying the time-series model.

6. The system according to claim 3, wherein the time-series model identified on the basis of the time-series data is an ARIMA model.

7. The system according to claim 1, wherein the fourth means is a server sizing means for deciding a configuration or setting of software or hardware of the computer on the basis of the resource utilization for each transaction estimated by the second means and the future amount of processing for each transaction predicted by the third means.

8. The system according to claim 1, wherein the fourth means is a server monitoring means for controlling monitor movement of the computer on the basis of the resource utilization for each transaction estimated by the first means and the future amount of the processing for each transaction predicted by the third means.

9. The system according to claim 8, wherein the server monitoring means further determines whether or not a warning should be given.

10. A computer implemented method for monitoring a server, comprising the steps of:

acquiring a first log containing resource utilization data and a second log containing transaction data from the server, and storing them in a storage device of the computer;

estimating resource utilization for each transaction on the basis of resource utilization data included in the first log and information on the number of transactions processed included in the second log;

performing a statistical time-series analysis on the time-series data of the number of executions for each transaction included in the second log, and thereby predicting a future number of executions for each transaction; and

predicting server resource demand on the basis of the resource utilization for each transaction estimated by the estimating step and the future amount of execution for each transaction predicted by the predicting step.

11. The method according to claim 10, wherein the estimating step comprises, obtaining resource utilization data at a plurality of times and the number of executions for each transaction at the corresponding time from the first log and the second log, and estimating the resource utilization for each transaction by a multiple linear regression analysis.

12. The method according to claim 10, wherein the predicting step comprises steps of:

identifying a time-series model by performing a statistical time-series analysis; and

predicting the number of executions for each transaction using the identified time-series model.

13. The means according to claim 12, wherein the identifying step comprises:

excluding a part of the time-series data and identifying the time-series model on the basis of the remaining data;

predicting the number of executions for each transaction corresponding to the excluded part of the time-series data by the identified time-series model; and

comparing a result of the prediction with the actual data of the excluded part to verify the time-series model.

14. A computer program product for causing a computer to monitor a server, the computer program product comprising:

first computer implemented instructions for acquiring a first log containing resource utilization data and a second log containing transaction data from a server and storing them in a storage device;

second computer implemented instructions for estimating resource utilization for each transaction concerned on the basis of resource utilization data of the entire server included in the first log and information on the number of executions of each transaction included in the second leg;

third computer implemented instructions for performing a statistical time-series analysis on the basis of time-series data of the number of executions of each transaction included in the second log, and thereby predicting an amount of execution for each transaction; and

fourth computer implemented instructions for predicting server resource utilization on the basis of the resource utilization for each transaction estimated by the second instructions and the amount of execution of each transaction predicted by the third instructions.

15. The computer program product according to claim 14, wherein the second computer implemented instructions comprises:

a computer implemented process for obtaining resource utilization at a plurality of times and the number of executions of each transaction at the corresponding time from the first log and the second log; and

a computer implemented process for estimating resource utilization for each transaction by a multiple linear regression analysis.

16. The computer program product according to claim 14, wherein the third computer implemented instructions comprises:

a computer implemented process for identifying a time-series model by a statistical time-series analysis; and

a computer implemented process for predicting the future amount of execution of each transaction using the identified time-series model.

17. The computer program product according to claim 16, wherein the process for identifying the time-series model comprises:

a computer implemented process for excluding a part of the time-series data and identifying the time-series model on the basis of the remaining data;

a computer implemented process for predicting the number of processing for each transaction corresponding to the excluded part of the time-series data by the identified time-series model; and

a computer implemented process for comparing a result of the prediction with the actual data of the excluded part to verify the time-series model.