To execute a load test, LoadView emulates virtual users that interact simultaneously with a target web resource, with each virtual user performing a sequence of actions specified by the device. In terms of LoadView, a Session is one completed run of the device by one virtual user.
In other words, each virtual user will be executing the device runs until the completion of the test. A sum of all runs that each virtual user will be able to do will give us the total number of sessions. Prior to the test execution, the system estimates the number of sessions by taking into account device duration time, overall test duration, the number of virtual users, and the load curve pattern.
The Estimated Number of Sessions for the Test
During the device validation process, the system calculates how long it takes to execute a single-user device run, known as Device Duration Time. Based on that value, and knowing the duration of the test and the number of virtual users, the system estimates the total number of completed sessions:
Estimated Sessions = VUM (Virtual User Minutes) / (Device Duration Time)
To put it another way, the Estimated Sessions value represents the number of sessions that can be executed under ideal conditions.
Why the Discrepancy in Estimated and Result Number of Sessions Occurs
Due to its nature, the Estimated Session value is preliminary and does not precisely identify a load test result. The actual number of completed sessions will differ depending on the fluctuations in server performance and response time, or errors that occurred while test execution.
For example, if a target server’s performance has dropped under the load, the response time may grow, and as a result, it may take more time to complete a single session. In this case, the session duration time will increase under the load and the actual number of sessions will be less than the estimated value.