Client-server architectures centralize the support for multi-user access to shared information and business processes. Part of any client-server design is an agreement between the project team and the customer (which may be the business itself) about how well the system should perform - the maximum expected response time for any query initiated by a user regardless of the number of other users interacting with that system at the same time. For a Web-based application, this "wait time" has been measured to be about 8 seconds; after 8 seconds, the typical user loses patience and browses elsewhere. For e-business websites there is no worse outcome.
IBM Rational Performance Tester enables teams to measure the response times of their client-server applications when subjected to multi-user loads. A built-in HTTP/HTTPS recorder and optional SQL recorder capture the communication initiated by client applications and the responses returned by the server environment. These scripts can then be customized and organized in a variety of ways to accurately reflect the habits of the various user profiles expected to use the application once it goes live. Using a variety of automated data correlation techniques, these tests can then be executed to reflect multiple, unique, concurrent actors - scaling to even thousands or tens of thousands of users. During execution, response times are collected and graphically represented, delivering real-time feedback about how well the system is performing. Once the load test has completed, multiple views to the response data can be analyzed to not only uncover system bottlenecks but to also dive into the transaction and look for root causes. With IBM Rational Performance Tester, teams are able to expose reliability problems in even the most complex systems before they go live, increasing the opportunity for defect capture and repair before it's too late.
