Getting Started

Getting Started

Load simulation against a Web application is usually implemented through a combination of hardware- and software-based emulation of users executing real-world transaction scenarios against an application. The hardware method usually requires several dedicated test clients and represents one user per client. This method can be quite expensive and time consuming to maintain, given the required hardware resources for heavy stress testing. The hardware method is neither the most effective nor the most efficient method of stress testing, but due to the nature of a given application, hardware-based stress testing maybe the only choice. With the software method, user transaction scenarios are typically recorded or captured by a software tool and converted to test scripts during a walkthrough of the application. When playing back or executing a test script, one machine typically acts as a controller to distribute the test script across multiple client machines. As the test runs, the controller produces load in unison to simulate many virtual users, and when the test is complete, it summarizes the test data from all of the clients. These virtual users create load on the backend servers and help to determine the stability and responsiveness of the Web application.

There are many software-based stress testing tools. While the purpose of these tools is to produce load by simulating many users, the syntax used for creating stress scripts and the load engine varies per tool. The stress tool discussed throughout this book is Microsoft Application Center Test (ACT). The goal of this chapter is not necessarily to make you an expert at using ACT but to give you a brief overview of the product and discuss the core concepts and features we frequently use. We will also demonstrate and explain how to create and verify dynamic test scripts using ACT to accurately simulate load against a .NET Web application

What is ACT?

Similar products