In LoadComplete, you can parameterize tests, that is, replace recorded values in requests with external or generated data.
LoadComplete simulates requests using data saved during the scenario recording. If you play back the recorded requests for multiple virtual users, users will send identical requests.
It can cause the following issues:
The simulated traffic will be unrealistic. In real life, users working with your web application are unlikely to use the same user names, passwords and other input values.
If you simulate many identical requests, data caching on the server will result in improbably high server performance. Test results will not show the real load the server would have in the production environment.
To make your virtual users behave like real-life users, you need to supply them with different parameter values. You do this by parameterizing requests, that is, by specifying various data for request parameters.
Prepare data you will use in your load tests. In LoadComplete, you can parameterize requests with –
Data loaded from external sources: Excel files, text files or databases.
Data that LoadComplete generates.
Create variables you will use to access the test data.
|Note:||You can replace a recorded value only with one variable. You cannot concatenate variable values.|
In LoadComplete, you can parameterize:
Form fields in HTML POST requests
Parameters the request body passes
Request header fields
Cookies (not recommended, for information on how to handle cookies in load tests, see Managing Cookies)
Hidden fields (usually, hidden fields pass helper information, have a specific format and are normally not parameterized)
Server URLs (see Changing the Tested Server)
|Note:||In LoadComplete, user scenarios can store up to 50 MB of data for a response. If the values you need are beyond this bound, they will not be parameterized.|
Often, the tested web server returns data that further requests use (for example, a user ticket or session ID). In LoadComplete, you can extract that data from server responses to pass it in further requests. To learn how to parameterize requests with data extracted from server responses, see Data Correlation.