Portable performance analytics system and method
Systems and apparatus for generating portable performance analytics for different applications across multiple diverse computing platforms and methods for making and using the same. In various embodiment, the system can compare results from multiple runs of a selected application under different virtualized environment conditions and correlate these results to a native baremetal performance of the application. The system thereby can identify scalability issues, noisy neighbor outliers or any other performance abnormalities in different run-time environments among other things.
1. A computer-implemented method for generating performance analytics for a selected software application across a plurality of diverse computing platforms, comprising:
executing the selected software application at each computing platform in the plurality of diverse computing platforms;
receiving one or more performance metrics from each computing platform in the plurality of diverse computing platforms about the execution of the selected software application at each computing platform in the plurality of diverse computing platforms;
comparing a performance of the selected software application at each one computing platform in the plurality of diverse computing platforms based on the one or more performance metrics;
detecting a weak point in the performance of the selected software application in a selected computing platform, wherein the selected computing platform is selected from the plurality of diverse computing platforms; and
performing a performance analysis for the selected software application.
2. The computer-implemented method of claim 1 , wherein said receiving the one or more performance metrics comprises measuring the one or more performance metrics of the selected software application at each computing platform in the plurality of diverse computing platforms, and further comprising evaluating the measured performance metrics.
3. The computer-implemented method of claim 2 , further comprising conducting performance characterization of the selected software application based upon said measuring the one or more performance metrics and said evaluating the measured performance metrics.
4. The computer-implemented method of claim 1 , wherein said detecting the weak point includes detecting a scalability issue for the selected software application, detecting a noisy neighbor outlier issue for the selected software application, detecting a bottleneck in the performance of the selected software application in the selected computing platform or a combination thereof.
5. The computer-implemented method of claim 1 , wherein said performing the performance analysis includes performing a price analysis for the selected software application.
6. The computer-implemented method of claim 1 , further comprising generating performance analytics for a second selected software application.
7. The computer-implemented method of claim 6 , wherein said generating the performance analytics for the second selected software application includes:
executing the second selected software application at each computing platform in the plurality of diverse computing platforms;
receiving one or more performance metrics from each computing platform in the plurality of diverse computing platforms about the execution of the second selected software application at each computing platform in the plurality of diverse computing platforms;
comparing a performance of the second selected software application at each one computing platform in the plurality of diverse computing platforms based on the one or more performance metrics from each computing platform in the plurality of diverse computing platforms about the execution of the second selected software application;
detecting a weak point in the performance of the second selected software application in a second selected computing platform, wherein the second selected computing platform is selected from the plurality of diverse computing platforms; and
performing a performance analysis for the second selected software application.
8. The computer-implemented method of claim 6 , wherein said generating the performance analytics for the second selected software application includes:
receiving one or more performance metrics from each computing platform in a second plurality of diverse computing platforms about the execution of the second selected software application at each computing platform in the second plurality of diverse computing platforms;
comparing a performance of the second selected software application at each computing platform in the second plurality of diverse computing platforms;
detecting a weak point in the performance of the second selected software application in a second selected computing platform among the second plurality of diverse computing platforms; and
performing a performance analysis for the second selected software application.
9. The computer-implemented method of claim 8 , wherein the plurality of diverse computing platforms and the second plurality of diverse computing platforms share at least one common computing platform.
10. The computer-implemented method of claim 6 , wherein said generating the performance analytics for the second selected software application is performed in parallel with generating the performance analytics for the selected software application.
11. The computer-implemented method of claim 6 , wherein said generating the performance analytics for the second selected software application is performed after generating the performance analytics for the selected software application.
12. The computer-implemented method of claim 1 , further comprising repeating said receiving the one or more performance metrics, said comparing the performance of the selected software application, said detecting the weak point and said performing the performance analysis for multiple runs of the selected software application.
13. The computer-implemented method of claim 1 , wherein said comparing the performance of the selected software application includes comparing the performance of the selected software application with historical performance of the selected software application.
14. The computer-implemented method of claim 13 , wherein said comparing the performance of the selected software application includes comparing the performance of the selected software application with historical performance of the selected software application of a predetermined computing platforms that is not among the plurality of diverse computing platforms.
15. The computer-implemented method of claim 1 , wherein said performing the performance analysis comprises performing the performance analysis for the selected software application in real time.
16. A non-transitory machine-readable storage media embodying a computer program product, said computer program product comprising:
instruction for executing a selected software application at each one of a plurality of diverse computing platforms
instruction for measuring one or more performance metrics of the selected software application at each one of the plurality of diverse computing platforms;
instruction for comparing a performance of the selected software application at each one of the plurality of diverse computing platforms based on the one or more performance metrics;
instruction for detecting a weak point in the performance of the selected software application in a selected computing platform, wherein the selected computing platform is one of the plurality of diverse computing platforms; and
instruction for performing a performance analysis for the selected software application.
17. The non-transitory machine-readable storage media of claim 16 , wherein said instruction for measuring, said instruction for comparing, said instruction for detecting and said instruction for performing each are encoded on the non-transitory machine-readable storage media.
18. A system for generating performance analytics for a selected software application, comprising:
a plurality of diverse computing platforms; and
a processing circuit for executing the selected software application at each one of the plurality of diverse computing platforms, measuring one or more performance metrics of the selected software application about the execution of the selected software application at each one of the plurality of diverse computing platforms, comparing a performance of the selected software application across each one of the plurality of diverse computing platforms based on the one or more performance metrics, and detecting a weak point in the performance of the selected software application in a selected diverse computing platform, wherein said processing circuit performs a performance analysis for the selected software application.
19. The system of claim 18 , wherein said diverse computing platforms include at least one supercomputing environment.
20. The system of claim 18 , wherein said diverse computing platforms include at least one cloud-based computing environment.