Granted Patent B2

US 8,635,498 · App. 13/123,702 · Granted Jan 21, 2014

Abstract

Claims (26)

1. A method for analyzing performance of an application, comprising:

collecting performance metric data from the application over time;

segmenting the performance metric data into time segments representing sets of contiguous time samples which exhibit similar performance metric behavior,

wherein each of the time segments relates to a period of different behavior within the performance metric data;

determining the presence of an anomaly in a time segment; and

correlating the anomalous segment with other data available to the system to determine the cause of the anomaly.

2. The method of claim 1 , wherein the performance metric data comprises at least one of response time, number of unavailabilities per unit of time, throughput, and hits per second.

3. The method of claim 1 , wherein determining the presence of an anomaly comprises comparing a baseline state of the application with the performance metric data from a time segment to determine the presence of an anomaly in the time segment.

4. The method of claim 1 , wherein determining the presence of an anomaly comprises comparing the performance metric data from a time segment with performance metric data from a past time segment to determine the presence of an anomaly in the time segment.

5. The method of claim 1 , further comprising querying a database regarding changes to the application or changes to a supporting component of the application that occurred during a time segment.

6. The method of claim 1 , further comprising querying a database to identify a support component of the application upon detection of an anomaly in the performance metric data from a time segment.

7. The method of claim 6 , wherein the other data available to the system for correlation includes a set of performance metric data of the support component from the time segment.

8. The method of claim 1 , further comprising performing a commonality analysis between a detected anomalous segment and past anomalous segments.

9. The method of claim 1 , further comprising identifying a pattern in the collected performance metric data, and classifying the anomaly into a category based on the identified pattern.

10. The method of claim 9 , further comprising projecting trends in the identified pattern to identify future anomalies.

11. A computer readable storage medium comprising a program for analyzing performance of an application, the program when executed by a computer processor causing the processor to perform the steps of:

collecting performance metric data from the application;

segmenting the performance metric data into time segments representing sets of contiguous time samples which exhibit similar performance metric behavior,

wherein each of the time segments relates to a period of different behavior within the performance metric data;

determining the presence of an anomaly in a time segment; and

correlating the performance metric data to determine the cause of the anomaly.

12. A system for analyzing performance of an application, comprising:

an analysis module configured to collect performance metric data from the application, the analysis module further configured to identify an anomaly in one or more of time segments of the collected performance metric data;

a segmentation engine configured to segment the collected performance metric data into time segments, representing sets of contiguous time samples which exhibit similar performance metric behavior,

wherein each of the time segments relates to a period of different behavior within the performance metric data; and

a correlation engine configured to correlate the time segments with other time-based data in the system to determine a cause of the identified anomaly.

Assignments (12)