How does LoadRunner work?

1. the controller is a central controller for managing and monitoring load tests. It starts numerous virtual users through the load generator (LG: load generator, send requests to applications to generate loads.

2. The real-time monitor captures cross-layer performance data, servers, and network resources and sends them back to the Controller.

3. The controller saves the data in the database, and the analyzer can start the analysis program to import the data.

Unlike winrunner, which can only simulate one client to use the system, LoadRunner can simulate thousands of virtual users.

The server Load balancer is controlled by vugen scripts that send non-user interface API calls that are the same as the protocols used during user operations. However, winrunner's graphical user interface allows a virtual user to simulate the keyboard hitting and clicking the mouse and other user interface operations. One machine can only run on one GUI, unless the LoadRunner Terminal Service Manager manages multiple machines that have started the Terminal Server Agent, and log on to a terminal service process.

During running, virtual users at the Thread level share a general memory pool, so multithreading enables each load generator to support more virtual users.

The status of virtual users on all load generators starts from "running" and passes through the init part of the script to the "ready" status, the virtual user is in the finished state of success or failure. When the load generator is overloaded, the virtual user will automatically "stop )".

Original from