Real-time database performance and availability monitoring method and system

• US 7,756,840 B2
• Filed: 11/03/2005
• Issued: 07/13/2010
• Est. Priority Date: 11/03/2004
• Status: Active Grant

• Current Assignee: Enteros, Inc.
• Original Assignee: DBA Infopower Incorporated
• Inventors: Warshawsky, Ron
• Primary Examiner(s): Trujillo; James
• Assistant Examiner(s): Moser; Bruce M
• Application Number:
  US11/265,244
• Publication Number:
  US 20060106851A1
• Time in Patent Office: 1,713 Days
• Field of Search: 707/101, 707/5, 707/102, 707/999.005, 707/999.101, 707/999.102, 707/999.05, 701/179
• US Class Current: 707/688
• CPC Class Codes: G06F 16/2358 Change logging, detection, and notification replication G06F16/27

Abstract

Database performance and availability monitoring of changes impacting database performance, availability and continuity to the underlying business may be performed. A method for doing so may include analytical and visual real-time analysis engines to identify and provide alert notifications on changes in database performance statistics (such as CPU consumption, physical I/O, etc.) related to a sample period of time on a single database or across multiple databases. Result data may be displayed through a series of charts and/or summary tables that may indicate whether correlations exist between unexpected database performance and relative changes in database performance statistical parameters.

Claims

1. A machine-accessible storage medium having computer-executable instructions that, upon execution by a processor, cause the processor to perform a method of database performance data processing comprising:

setting a value of a time window size TP, where TP is a time period between a current moment of time t and a moment of time TP periods ago, t−

TP+1;

obtaining a value of a database performance statistic of at least one database in the current moment of time t as VAL(t);

determining a highest value of the database performance statistic of the at least one database within a time period from t−

TP+1 to t, as HV(t);

determining a lowest value of the database performance statistic of the at least one database within the time period from t−

TP+1 to t, as LV(t); and

calculating a normalized value of a data point at time t of the at least one database as (VAL(t)−

LV(t))/(HV(t)−

LV(t));

determining the performance of the at least one database based on the calculated normalized value.

2. The medium of claim 1, wherein said obtaining a value of a database performance statistic comprises:

choosing VAL(t) from a database performance statistic data set, DS, comprising data elements DS(t), wherein a data element value DS(t) at time t is calculated as D(t)−

D(t−

1), where D(t) is a value of a database performance statistic at time t and D(t−

1) is a value of the database performance statistic at a previous time, t−

1;

wherein said highest value and said lowest value are determined based on said data set, DS; and

wherein said normalized value at time t is calculated based on the data set DS, using VAL(t) and said highest and lowest values.

3. The medium of claim 2, having further computer-executable instructions that, upon execution by a processor, cause the processor to perform a method that further comprises:

identifying sets of data points from a set of single database performance statistics for the at least one database, wherein a corresponding time stamp value is recorded for each data point; and

displaying the data points of at least one of said sets of data points on a chart, said chart comprising a data set display area, a vertical database performance statistic value axis, and a horizontal axis of time.

4. The medium of claim 2, wherein said normalized value is compared with a threshold value selected from the group consisting of:

a user defined threshold value and a default threshold value.

5. The medium of claim 1, wherein said obtaining a value of a database performance statistic comprises:

setting VAL(t) equal to a function of at least one database performance statistic, wherein said highest value, said lowest value, and said normalized value are determined based on the same function of at least one database performance statistic.

6. The medium of claim 5, wherein said function comprises a moving average of a database performance statistic.

7. The medium of claim 5, having further computer-executable instructions that, upon execution by a processor, cause the processor to perform a method that further comprises:

identifying sets of data points from a set of single database performance statistics for the at least one database, wherein a corresponding time stamp value is recorded for each data point; and

displaying the data points of at least one of said sets of data points on a chart, said chart comprising a data set display area, a vertical database performance statistic value axis, and a horizontal axis of time.

8. The medium according to claim 6, further comprising computer-executable instructions such that the method further comprises:

obtaining from a user a period for taking said moving average of a database performance statistic, thereby obtaining a moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values;

wherein said moving average uses said moving average period.

9. The medium according to claim 6, further comprising computer-executable instructions such that the method further comprises:

computing a database performance statistic moving average period, defining a subset of values over which the moving average is to be taken, as a user defined fraction of an overall time span of underlying data values of a database parameter data set;

wherein said moving average uses said moving average period.

10. The medium of claim 5, wherein said normalized value is compared with a threshold value selected from the group consisting of:

a user defined threshold value and a default threshold value.

11. The medium of claim 1, having further computer-executable instructions that, upon execution by a processor, cause the processor to perform a method that further comprises:

identifying sets of data points from a set of single database performance statistics for the at least one database, wherein a corresponding time stamp value is recorded for each data point; and

displaying the data points of at least one of said sets of data points on a chart, said chart comprising a data set display area, a vertical database performance statistic value axis, and a horizontal axis of time.

12. The medium of claim 11, wherein said sets of data points correspond to a single similar statistic across the at least one database, and wherein said sets of data points are displayed on a single chart.

13. The medium according to claim 12, further comprising computer-executable instructions such that the method further comprises:

obtaining from a user a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values; and

calculating a database performance statistic moving average data set based on the sets of data points using said moving average period.

14. The medium of claim 11, wherein said sets of data points are displayed on a single chart for a single similar statistic for a logical database cluster, wherein said logical database cluster comprises two or more physical databases.

15. The medium according to claim 14, further comprising computer-executable instructions such that the method further comprises:

obtaining from a user a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values; and

calculating a database performance statistic moving average data set based on the sets of data points using said moving average period.

16. The medium of claim 11, wherein said sets of data points are displayed on a single chart for the at least one database performance statistics for a single physical database.

17. The medium according to claim 16, further comprising computer-executable instructions such that the method further comprises:

obtaining from a user a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the movie age is to be taken within an overall set of values; and

calculating a database performance statistic moving average data set based on the sets of data points using said moving average period.

18. The medium of claim 11, wherein said sets of data points are displayed on a single chart for multiple database performance statistics for a logical database cluster, wherein said logical database cluster comprises two or more physical databases.

19. The medium according to claim 18, further comprising computer-executable instructions such that the method further comprises:

obtaining from a user a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values; and

calculating a database performance statistic moving average data set based on the sets of data points using said moving average period.

20. The medium of claim 11, further comprising additional computer-executable instructions such that the method further comprises:

calculating at least one set of data points on a same database performance statistic using data from historical time periods to thus obtain one or more sets of historical data points; and

displaying said one or more sets of historical data points on said chart, along with said one or more sets of data points.

21. The medium according to claim 11, further comprising computer-executable instructions such that the method further comprises:

obtaining from a user a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values; and

calculating a database performance statistic moving average data set based on the sets of data points using said moving average period.

22. The medium of claim 11, wherein said one or more database performance statistics are based on SQL statistics associated with one or more SQL statements executed in one or more databases.

23. The medium of claim 11, wherein said one or more database parameters are related to SQL statistics associated with one or more SQL statements executed in one or more databases.

24. The medium of claim 1, wherein said normalized value is compared with a threshold value selected from the group consisting of:

a user defined threshold value and a default threshold value.

25. The medium of claim 1, wherein said obtaining a value of a database performance statistic comprises:

receiving a value of a database performance statistic from a database monitoring agent.

26. A machine-accessible storage medium having computer-executable instructions that, upon execution by a processor, cause the processor to perform a method of database performance analysis comprising:

calculating a database performance statistic moving average of at least one database based on a database performance statistic using a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values, wherein said database performance statistic moving average period is one of the group consisting of;

a user-defined period and a predetermined default period; and

comparing a value of the database performance statistic with the value of the moving average of the at least one database to identify a percent difference between the value of the database performance statistic and the value of the moving average;

wherein said database performance statistic is selected from the group consisting of;

a direct database performance statistic and a derived database performance statistic;

determining the performance of the at least one database based on the identified percent difference.

27. The medium of claim 26, wherein said method further comprises:

obtaining one or more values of said database performance statistic from at least one database monitoring agent.

28. A machine-accessible storage medium having computer-executable instructions that, upon execution by a processor, cause the processor to perform a method of database performance analysis comprising:

comparing a current value of a database performance statistic of at least one database corresponding to a current monitoring time period during which the database performance statistic was monitored with a previous value of the same database performance statistic obtained during a previous monitoring time period during which the database performance statistic was monitored to identify a percent difference between said current value and said previous value;

wherein said database performance statistic is selected from the group consisting of;

a direct database performance statistic and a derived database performance statistic

determining the performance of the at least one database based on the identified percent difference.

29. The medium of claim 28, wherein said method further comprises:

obtaining one or more values of said database performance statistic from at least one database monitoring agent.

30. A method of database performance data processing comprising:

setting a value of a time window size TP, where TP is a time period between a current moment of time t and a moment of time TP periods ago, t−

TP+1;

obtaining a value of a database performance statistic of at least one database in the current moment of time t as VAL(t);

determining a highest value of the database performance statistic of the at least one database within a time period from t−

TP+1 to t, as HV(t);

determining a lowest value of the database performance statistic of the at least one database within the time period from t−

TP+1 to t, as LV(t); and

calculating a normalized value of a data point at time t of the at least one database as (VAL(t)−

LV(t))/(HV(t)−

LV(t));

determining the performance of the at least one database based on the calculated normalized value.

31. The method of claim 30, wherein said obtaining a value of a database performance statistic comprises:

choosing VAL(t) from a database performance statistic data set, DS, comprising data elements DS(t), wherein a data element value DS(t) at time t is calculated as D(t)−

D(t−

1), where D(t) is a value of a database performance statistic at time t and D(t−

1) is a value of the database performance statistic at a previous time, t−

1;

wherein said highest value and said lowest value are determined based on said data set, DS; and

wherein said normalized value at time t is calculated based on the data set DS, using VAL(t) and said highest and lowest values.

32. The method of claim 30, wherein said obtaining a value of a database performance statistic comprises:

setting VAL(t) equal to a function of at least one database performance statistic, wherein said highest value, said lowest value, and said normalized value are determined based on the same function of at least one database performance statistic.

33. The method of claim 30, further comprising:

identifying sets of data points from a set of single database performance statistics for a single database, wherein a corresponding time stamp value is recorded for each data point; and

displaying the data points of at least one of said sets of data points on a chart, said chart comprising a data set display area, a vertical database performance statistic value axis, and a horizontal axis of time.

34. The method of claim 30, wherein said normalized value is compared with a threshold value selected from the group consisting of:

a user defined threshold value and a default threshold value.

35. The method of claim 30, wherein said obtaining a value of a database performance statistic comprises:

receiving a value of a database performance statistic from a database monitoring agent.

36. The method of claim 30, further comprising:

downloading computer-executable instructions that, upon execution by a processor, cause the processor to perform said setting a value, said obtaining a value of a database performance statistic, said determining a highest value, said determining a lowest value, and said calculating a normalized value.

37. A method of database performance analysis comprising:

calculating a database performance statistic moving average of at least one database based on a database performance statistic using a database performance statistic moving average period that defines a subset of values of the database performance statistic over which the moving average is to be taken within an overall set of values, wherein said database performance statistic moving average period is one of the group consisting of;

a user-defined period and a predetermined default period; and

comparing a value of the database performance statistic with the value of the moving average of the at least one database to identify a percent difference between the value of the database performance statistic and the value of the moving average;

wherein said database performance statistic is selected from the group consisting of;

a direct database performance statistic and a derived database performance statistic;

determining the performance of the at least one database based on the identified percent difference.

38. The method of claim 37, further comprising:

obtaining one or more values of said database performance statistic from at least one database monitoring agent.

39. The method of claim 37 further comprising:

downloading computer-executable instructions that, upon execution by a processor, cause the processor to perform said calculating and said comparing.

40. A method of database performance analysis comprising:

comparing a current value of a database performance statistic of at least one database corresponding to a current monitoring time period during which the database performance statistic was monitored with a previous value of the same database performance statistic obtained during a previous monitoring time period during which the database performance statistic was monitored to identify a percent difference between said current value and said previous value;

wherein said database performance statistic is selected from the group consisting of;

a direct database performance statistic and a derived database performance statistic

determining the performance of the at least one database based on the identified percent difference.

41. The method of claim 40, further comprising:

obtaining one or more values of said database performance statistic from at least one database monitoring agent.

42. The method of claim 40 further comprising:

downloading computer-executable instructions that, upon execution by a processor, cause the processor to perform said calculating and said comparing.