Virtual users have exited prior to stage completion. At the end of stage %1 there were %2 users running when %3 were expected. A common reason for this is a schedule which has assigned an insufficient amount of work (for one or more User Groups), to keep the associated virtual users active for the full amount of time specified by the stage duration(s). Review the schedule for sufficient workload. When using multiple stages it is strongly recommended to use infinite loops to ensure that virtual users always have enough work scheduled regardless of the total duration of all the stages. Another possible reason is that one or more virtual users exited upon encountering a serious error; consult the Test Log or Common Base Event XML log on the agent for more information. Also see the Troubleshooting section of the online help.
During schedule execution, at the end of the current stage, the actual number of users running did not match the expected number of users. For example, if the current stage specifies that 100 users should run for 1 hour and only 90 users are running at the end of the hour, this message is displayed.
Check the Error Log for further information on the error. To open the Error Log, click Window > Show View > Error Log.
Typically, this message is displayed when virtual users did not have enough work to do for the duration of the stage. For schedules that contain more than one stage, verify that the workload under each user group is contained inside an infinite loop. Use infinite loops because the stage duration is controlled by the time when users stop.
If virtual users have sufficient workload, look in the test log for more information about why virtual users stopped. The virtual users that stopped might have encountered errors.
By default, this message is displayed when the number of expected users does not match the number of actual users running at the end of a stage. You can change this setting to specify the percentage of users that may stop during a stage without being considered an error. To change the error condition, create the -DrptStopTolerance property in the eclipse.ini file in the installation directory. For example, -DrptStopTolerance=80 specifies that 80% of the users may stop unexpectedly during stage execution without being considered an error.