System and method for extracting and forecasting computing resource data such as CPU consumption using autoregressive methodology

• US 6,574,587 B2
• Filed: 02/27/1998
• Issued: 06/03/2003
• Est. Priority Date: 02/27/1998
• Status: Expired due to Fees

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: MCI Communications Corporation (Verizon Communications Inc.)
• Inventors: Waclawski, Anthony C.
• Primary Examiner(s): KEMPER, MELANIE A
• Application Number:
  US09/031,966
• Publication Number:
  US 20010013008A1
• Time in Patent Office: 1,922 Days
• Field of Search: 705/1, 709/104, 709/103, 709/224, 709/226, 714/1, 714/4, 702/179, 702/181, 702/182, 702/186
• US Class Current: 702/186
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/3423 where the assessed time is active or idle time
• 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 2201/835 Timestamp
• CPC Class Codes: G06F 2201/875 Monitoring of systems including the internet

Abstract

A system and method for extracting and forecasting computing resource data such as workload consumption of mainframe computing resources using an autoregressive model. The system and method forecast mainframe central processing unit (CPU) consumption with ninety-five percent accuracy using historical performance data. The system and method also provide an upper ninety-five percent confidence level and a lower ninety-five percent confidence level. The system and method retrieve performance records from a computer platform in one second intervals, statistically collapses the one second performance data into fifteen minute performance data, statistically collapses the fifteen minute performance data into one week performance data, and generates a time series equivalent to collecting performance data at one week intervals. The system and method ensure that the resulting time series is statistically stationary, and applies an autoregressive construct to the time series to generate forecast of future CPU utilization, as well as to generate reports and graphs comparing actual vs. forecast CPU utilization. Because the system and method rely on electronically generated empirical historical computer performance data as an input, they provide a turnkey solution to CPU consumption forecasting that can be implemented easily by any system network manager.

Claims

1. In a computing platform having a plurality of resources, each resource includes at least one aspect, a method for forecasting at least one aspect of the plurality of resources to produce computing platform resource performance forecasts, the method comprising the steps of:

collecting at intervals performance data associated with a performance metric from a computing platform resource;

statistically collapsing the collected performance data associated with a performance metric at least once to produce a time series and storing the time series;

determining whether the time series is statistically stationary using a plurality of chi-square tests;

converting the time series to a statistically stationary time series when the time series is determined not to be statistically stationary;

bypassing the converting when the time series is determined to be statistically stationary;

adding a date/time stamp to the time series, including converting the time series date/time stamp to a number of seconds equivalent to the value represented by the date/time stamp;

applying an autoregressive modeling tool to the time series to produce computing platform resource performance forecasts; and

outputting the resource performance forecasts in a format such that computing platform resources can be modified based on the data.

12. The computing platform according to claim 1, further comprising feeding actual future performance data into the autoregressive modeling tool to generate recursive forecasts.

13. The method of claim 1, wherein the statistically collapsing the collected performance data associated with a performance metric at least once comprises:

statistically collapsing the collected performance data to produce a first time series,

storing the first time series in a database,

extracting the first time series from the database, and

statistically collapsing the first time series to produce a second time series; and

wherein the determining whether the time series is statistically stationary comprises;

determining whether the second time series is statistically stationary.

2. A system to forecast performance of at least one computing platform resource, comprising:

a resource manager for collecting performance data associated with a performance metric from a computing platform resource;

a first statistical collapser receiving input from the resource manager, the first statistical collapser collapsing the data received from the resource manager into a first time series;

a second statistical collapser, the second statistical collapser collapsing the data received from the first statistical collapser into a second time series;

a time series analyzer coupled to the second statistical collapser and configured to;

determine whether the second time series is statistically stationary using a plurality of chi-square tests,

convert the second time series to a statistically stationary time series when the second time series is determined not to be statistically stationary, and

bypass the converting when the second time series is determined to be statistically stationary;

a time point converter coupled to the time series analyzer, the time point converter being operable to add a date/time stamp to the second time series and wherein the time point converter converts the time series date/time stamp to a number of seconds equivalent to the value represented by the date/time stamp;

an autoregressive modeling tool coupled to the time series analyzer, the autoregressive modeling tool producing computing platform resource performance forecasts; and

a result processor coupled to the autoregressive modeling tool.

3. The system according to claim 2, further comprising a database coupled to the first statistical collapser.

4. The system according to claim 3, further comprising a data extractor coupled to the database.

5. The system according to claim 2, wherein the performance metric represents one of redundancy or efficiency of at least one computing platform service.

6. The system according to claim 2, wherein the performance metric represents utilization and the computing platform resource comprises a central processing unit.

7. The system according to claim 2, wherein the first statistical collapser comprises a fifteen minute time series generator.

8. The system according to claim 2, wherein the second statistical collapser comprises a weekly time series generator.

9. The system according to claim 2, wherein the time series analyzer comprises a time series differencer.

10. The system according to claim 2, wherein the autoregressive modeling tool comprises an autoregressive moving average modeling tool.

11. The system according to claim 2, wherein the autoregressive modeling tool comprises an autoregressive integrated moving average modeling tool.