kkyv wrote:What kind of data do you typically extract from each test playback?
The most common data we extract is what we call the "split times" in each test playback.
It's easy to put
SaveRecord commands in the script to record the elapsec time since script playback for a sequence of events in the test. You might think of two or three or four key values that tell you, after the fact, whether that particular test playback was a slow performer or a fast performer.
In LoadTest mode this works the same way, but normally we have each Browser User (BU) write to its own file. After the LoadTest run is done we accumulate all of the data from all of the BUs and process it (usuallly with a spread sheet) to get the overall min/average/maximum times for each defined step.
The eValid Team