Method and system for processing application performance data ouside of monitored applications to limit overhead caused by monitoring

• US 20080288212A1
• Filed: 05/15/2008
• Published: 11/20/2008
• Est. Priority Date: 05/15/2007
• Status: Active Grant

• Current Assignee: Dynatrace LLC (Dynatrace, Inc.)
• Original Assignee: dynaTrace Software GmbH (Dynatrace, Inc.)
• Inventors: Greifeneder, Bernd
• Granted Patent: US 7,957,934 B2
• Time in Patent Office: Days
• US Class Current: 702/182
• CPC Class Codes: G06F 11/3466 Performance evaluation by tracing or monitoring
• CPC Class Codes: G06F 2201/86 Event-based monitoring
• CPC Class Codes: G06F 2201/865 Monitoring of software
• CPC Class Codes: G06F 2201/88 Monitoring involving counting

Abstract

The present invention provides a system and method that aggregates and processes monitoring data acquired inside a monitored application outside of the said application to reduce overhead caused by monitoring tasks in the monitored application. Monitoring data acquired by monitoring sensors is stored in a ring buffer after acquisition, which decouples acquisition of data and its transfer, enabling asynchronous sending of performance data to a monitoring server for analysis. Memory consumption of said ring buffer is limited to a fixed size which keeps memory consumption caused by monitoring processes low and predictable. Correlation, analysis, all computation of monitoring data, and bytecode manipulation to place sensors is performed outside of the monitored application, by a monitoring server.

Claims

1. A method to process application performance data, comprising the steps of:

injecting an agent into a monitored application;

capturing a byte code of the monitored application by the agent;

sending the captured byte code of the monitored application to a monitoring server external to the monitored application;

generating instrumented byte code from the captured byte code;

providing the instrumented byte code to the monitored application;

loading the instrumented byte code by the monitored application;

generating measurement data from the instrumented byte code;

writing the measurement data to a data set of fixed size.

2. A method to process application performance data as in claim 1, wherein the measurement data include a measurement value.

3. A method to process application performance data as in claim 1, wherein the measurement data includes a unique ID.

4. A method to process application performance data as in claim 1, wherein the measurement data include a measurement value and a unique ID.

5. A method to process application performance data as in claim 4, wherein the measurement values and the unique ID are combined to obtain a performance measurement.

6. A method to process application performance data as in claim 1, wherein the data set of fixed size includes a ring buffer.

7. A method to process application performance data as in claim 1, wherein the data set is adapted to be asynchronous read out.

8. A method to process application performance data as in claim 2, wherein the sensor data includes counting sensor data.

9. A method to process application performance data as in claim 2, wherein the sensor data includes timing sensor data.

10. A method to process application performance data as in claim 1, wherein the agent captures the byte code by interrupting the loading of the byte code.

11. A method to process application performance data as in claim 1, wherein the method includes the step of extracting metadata from the byte code.

12. A method to process application performance data as in claim 11, wherein the metadata is used to select a method.

13. As in claim 12, wherein the method includes a comparing method.

14. A method to process application performance data as in claim 1, wherein the captured byte code is external to the monitored application;

and

wherein the method includes the steps of generating metadata describing the instrumented byte code external to the monitored application and storing said metadata external to the monitored application;

applying a key to the instrumented byte code which identifies the metadata describing the instrumented byte code; and

correlating measurement data generated by a specific portion of instrumented byte code with the metadata describing the specific portion of instrumented byte code outside of the monitored application.

15. A method to process application performance data as in claim 1, wherein an event record is placed within the database of fixed size.

16. A method to process application performance data as in claim 15, wherein the monitoring server collects the event record.

17. A method to process application performance data as in claim 5, wherein the monitoring server compares the sensor ID of the event record with stored sensor ID in order to determine if the event is to be measured.

18. A method to process application performance data as in claim 17, wherein the event record is discarded if the sensor ID of the event record does not match the stored sensor ID.

19. A method to process application performance data as in claim 18, wherein the event record is combined with stored sensor metadata if the sensor ID of the event record matches the stored sensor ID.

20. A method to process application performance data as in claim 1, wherein the method includes the step of bypassing placing the measurement data into the data set if the capacity of the data set has been exceeded.