The most important easurement

[tomz]

From the client side, what do you think is the most important easurement?

[eValid]

From the client side, what do you think is the most important measurement?

You didn't say if this was for functional/regression testing, for monitoring, or for server loading. The answer is slightly different for each:

Functional/Regression Testing: The basis fact you want is whether the script ran to the end without any problems, a fact that indicates "no regression" in the underlying application. So the metric used is "error count."

Monitoring: Typical monitoring tests are very stable, and the main question is "how long did this playback take?" Of course, if the application is off the air then the playback will time out and you'll see an ERROR status; this also can be indicated when there is a maximum allowable time, sensed in eValid playback with a timer check command. So here the main metric used is "time".

Server Loading: The direct response time for a section of a playback will grow as the total load on the server grows, so for server loading the most direct metric is intra-script playback elapsed time. In typical loading scenarios as the load increases you are also going to be looking at the server stack's performance, using the performance monitoring facilities on those machines. In the first case the metric is "time" and in the second case the metric is "percent of capacity."

eValid Support