Support

Are you new to load testing, or it's been a while? Learn about a few of the goals, motivations, and benefits to load testing your web application, and when it makes sense (and might not make sense) to invest the time and effort.

No test or simulation can entirely eliminate the risk of site failure, but if you employ a few load testing best practices, you can reduce the risk of crashing in a high-traffic event by 90% or even 99%... without too much effort.

Get acquainted with important terms and definitions used in the Loadster manual and documentation, and learn about the high-level process of load testing with Loadster, from scripting to load testing to analyzing the results.

Loadster's protocol scripts automate at the HTTP protocol layer, firing off a sequence of requests and interpreting the responses that come back. This type of scripting is ideal for testing HTTP APIs (REST, SOAP, etc) and also for cost-effectively load testing simple static websites.

Loadster's browser scripts control real headless web browsers, and they're the easiest way to load test complicated web applications. Load testing with real browsers greatly simplifies OAuth/SAML flows, video streaming, WebSockets, and many other things that used to be tricky to load test.

Particularly when testing dynamic web applications, a realistic load test will need to make different requests for each simulated user. Use variables and expressions for storing and reusing values, generating dynamic values, and encoding or decoding data.

Use Loadster's dynamic datasets to supply each simulated user in your load test with unique data. Generate test data or enter it manually with the editor, and then "bind" the dataset to a variable in your script, so each bot pulls values from it.

If you need to evaluate conditional logic, loop through values, or parse a complicated data structure in your load test scripts, take a look at code blocks. Anything you can do with ordinary steps can be done in JavaScript, and then some.

Ready to run a load test with multiple concurrent bots or virtual users? Here's how to structure the groups of bots in your test, and what to look for l while the test is running, so you can tell if the test is going according to plan.