Automatic performance statistical comparison between two periods

• US 20070083564A1
• Filed: 10/07/2005
• Published: 04/12/2007
• Est. Priority Date: 10/07/2005
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Ramacher, Mark, Wood, Graham, Green, Russell, Green, Konstantina, Grant, Cecilia
• Granted Patent: US 7,526,409 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/217 Database tuning G06F16/2282 takes precedence; database performance monitoring G06F11/3409
• CPC Class Codes: G06Q 10/0639 Performance analysis of employees; Performance analysis of enterprise or organisation operations

Abstract

A technique for automatically generating a report comprising normalized differences in performance statistics between two separate periods. In one embodiment of the invention, database performance statistics are collected on a periodic basis over various time periods. In order to accurately compare database system performance between two discrete periods of time, the difference in the performance statistics of each period are normalized prior to comparing the two periods with each other. By normalizing the statistical differences in each period prior to comparing the differences, periods of different lengths of time as well as different levels of database system load may be compared. In one embodiment, a report is automatically generated which lists the performance statistics being evaluated, the difference in the statistic between each period, the value of each statistical difference as normalized by database time, and the difference between the normalized values.

Claims

1. A method of representing database performance characteristics, the method comprising:

collecting at least two data values representing measurements of database performance, wherein the at least two data values correspond to two different periods of time;

for each of the two periods, determining a database time that represents an amount of time a database system works on calls by users to the database system; and

normalizing the at least two data values based on the respective database time for the respective periods of time.

2. The method of claim 1, further comprising automatically generating a report based on the normalized values.

3. The method of claim 2, wherein the report is ordered by amount of difference between the normalized data values.

4. The method of claim 2, wherein the report values are generated as the result of a SQL query.

5. The method of claim 1, wherein the at least two data values include cumulative values.

6. The method of claim 1, wherein the first of the at least two data values comprises:

the difference between statistical values captured by a snapshot at the beginning of a first period and statistical values captured by a snapshot at the end of the first period, and

the second of the at least two data values comprises;

the difference between statistical values captured by a snapshot at the beginning of a second period and statistical values captured by a snapshot at the end of the second period.

7. The method of claim 1, wherein the measurements of database performance comprise buffer busy waits.

8. The method of claim 1, wherein the measurements of database performance comprise log file syncs.

9. The method of claim 1, wherein the measurements of database performance comprise user commits.

10. The method of claim 1, wherein the measurements of database performance comprise session logical reads.