After configuring and starting a test, VPT will begin to collect response data from the target application and display aggregated metrics describing how the application is handling the workload. The data is shown in near-real time so that you can see how you application is handling load during the test and make decisions quickly to reduce cost and increase efficiency.

If you assigned any APM profiles to the test run, you’ll see an additional chart that displays the profiles’ selected metrics at the bottom of the page.

While the test is running, you will see the number of virtual users, hits per second, and errors graphed over the duration of the test. You will also see numerical metrics for transactions per second, average response time, and errors, aggregated for the entire test. If the target application responds with errors, they will be displayed on the page so you can better understand the situation.

There may be times when your test experiences an abnormal number of errors, which can be caused by a misconfiguration, network issue, or problem with the target application. In this case, instead of waiting for the test to end naturally, you can stop it manually before it reaches its configured duration. There are two ways to end the test manually.

Cancelling a test will end the test, but it will allow the virtual users to finish their current action before stopping. This is a great option if you don’t need the test to end immediately or if there could be inconsistencies because of disconnections or incomplete submissions.

Killing a test will cause the test to end more abruptly. When the test is killed, the virtual users are stopped as soon as the signal is received. This can lead to disconnections and interrupts. This is ideal if the test is bringing down applications or servers.

Regardless of how the test ends, all the data generated will be visible on the report page upon test completion.