One of my Web Service test suites has 174 test cases and it takes QTP... 7 minutes to execute them all. At first it may seems to be a resonable amount of time, especially compared to simular GUI operations which would have taken hours in test execution.
When I execute the same test suite running as a vbs-file instead I'm down to 1 minute and 45 seconds!
What are the differences? The code is the same apart from that all QTP specific log functions like Reporter are replaced by custom logging to files (which should be slower since 1 folder and 2 xml-files are created for each test case).
My conclusion is that QTP has a lot of overhead features which affects test execution performance in a negative way, especially considering that must features are not that useful in the long run (at least not for me).
Ps. My QTP settings are Run mode "fast" and no QTP recovery scenario is involved
When I execute the same test suite running as a vbs-file instead I'm down to 1 minute and 45 seconds!
What are the differences? The code is the same apart from that all QTP specific log functions like Reporter are replaced by custom logging to files (which should be slower since 1 folder and 2 xml-files are created for each test case).
My conclusion is that QTP has a lot of overhead features which affects test execution performance in a negative way, especially considering that must features are not that useful in the long run (at least not for me).
Ps. My QTP settings are Run mode "fast" and no QTP recovery scenario is involved
No comments:
Post a Comment