eddiem wrote:How does the eValid footprint grow over time if you run a test multiple times?
You're right. The footprint (the total amount of allocated memory) does grow over time. We don't like that, but there is nothing that can be done about it.
All browsers' footprints grow over time, IE and eValid pretty much equally.
Clearing the cache and cookies doesn't reduce the footprint size. Killing the eValid process and restarting does start it all over.
In LoadTest mode there is a "ReSpawn" parameter that forces the process to die and self-restart every specified number of playbacks. This keeps the footprint small, but at the expense of having to reload the entire eValid executable each time you do it. We typically will set the ReSpawn value at about 10 repetitions. That seems to be a good balance between those two factors.
All of these things taken into account, what we have found is that footprint growth doesn't really seem to get in the way of delivering realistic user perspective timing/performance data!