07-17-2015 4:35 PM
Hello,
I have created a simple parameterized test using eCATT, a test data container that contains some values for the parameters and a test configuration the refers to these variants. I want to schedule an automated "sweep" through all of the variants to compare the performance of the variants to each other and watch the performance changes over time. I have succeeded in executing the test configuration (manually, not yet scheduled) from ST30, and I can see that the system is processing all of the variants. In the statistics I can extract from ST30/ST33, I can only see data referring to the entire test configuration, not the individual variants. Is it possible to somehow add the variant information to the statistics? If not, what other options (besides logging the runtime with ABAP statements manually and storing it in a custom table) do I have?
Thanks in advance
Volker Wegert
07-24-2015 3:46 PM
Hello Volker,
I have some ideas, that maybe help.
I don't know how you executed all variants in one configuration, but if you cascade different eCatt scripts and call them in one Metascript using the REF command this could be a way.
So create one master eCatt script that you embed into the eCatt testconfiguration.
Call there as many subscripts ( using REF command ) and each Sub eCatt script uses another
variant.
Then use the PERF ENDPERF command around each REF call in the meta script.
you can Label each PERF ENDPERF with a string. In that case take the Name of the variant that is used in the script called by the REF command.
Then start the testconf in ST30 and use the Checkbox modular measurement.
Only then the PERF ENDPERF is considered. BTW: this is normally designed to measure only parts of an eCatt script 😉
Now the measurment results are aggregated for each different PERF ENDPERF Label.
Remark: if you Label different PERF ENDPERF commands with identical Labels they will be aggregated together in one result.
This could be the solution for your requirement I hope.
Best Regards
Armin Hechler-Stark
07-24-2015 3:46 PM
Hello Volker,
I have some ideas, that maybe help.
I don't know how you executed all variants in one configuration, but if you cascade different eCatt scripts and call them in one Metascript using the REF command this could be a way.
So create one master eCatt script that you embed into the eCatt testconfiguration.
Call there as many subscripts ( using REF command ) and each Sub eCatt script uses another
variant.
Then use the PERF ENDPERF command around each REF call in the meta script.
you can Label each PERF ENDPERF with a string. In that case take the Name of the variant that is used in the script called by the REF command.
Then start the testconf in ST30 and use the Checkbox modular measurement.
Only then the PERF ENDPERF is considered. BTW: this is normally designed to measure only parts of an eCatt script 😉
Now the measurment results are aggregated for each different PERF ENDPERF Label.
Remark: if you Label different PERF ENDPERF commands with identical Labels they will be aggregated together in one result.
This could be the solution for your requirement I hope.
Best Regards
Armin Hechler-Stark
10-27-2015 12:14 PM
Armin,
thank you for your time and sorry for the late response. In the meantime, I have switched to a slightly different approach since I wanted to run automated tests using the Frontend Scheduler and didn't find a way to automate the ST30 without some serious front-end (non-SAP) UI scripting. At the moment, I'm using a custom table to do the logging, and among all of the parts that are giving me a headache in this project, at least this isn't one...
Best regards
Volker Wegert