Applies to ReadyAPI 2.7, last modified on May 14, 2019

Load Testing means running many functional tests in parallel or in series to simulate a load on a server and see how it behaves under this load. To run each functional test, ReadyAPI uses a virtual user (such users are called virtual, because they do not exist). The number of virtual users to be simulated is not always large, it depends on what the tested server calls the "expected load" and on the test strategy you choose. For more information about using different load strategies, see Testing Scenarios.

A good load test does not send random requests to a server. Instead, it simulates user behavior. For example, a web site provides API. You want to send API requests for it to handle, and then check whether it responds correctly and quickly enough.

Good load tests also assert performance to provide insight into the reasons for slow server responses. Because of the way servers operate, there are many possible causes of test slowdowns or sudden spikes of poor performance. For more information about asserting performance in LoadUI, see About Assertions.

Sometimes the issue occurs on the server side. A good load testing tool should be able to access all server-side metrics and display them to the tester during the test run, and store the results. For more information about monitoring servers in LoadUI, see About Server Monitoring.

See Also

Tutorials and Samples
Setting Up Load Tests
Distributed Load Testing

Highlight search results