Socket API performance testing overview

With IBM® Rational® Performance Tester Extension for Socket Protocols, you can test the performance of any application that uses a TCP/IP socket-based protocol.

Informative performance test results rely upon sound test development. Each of the following stages contributes to the generation of meaningful test results:
  • Test creation. You create your test by recording a session with a client application. Typically, the recorded session starts when you run the client application. You then interact with the application in order to produce relevant network traffic, and the session ends when you close the application or end the recording. The recording is used to generate a performance test that reproduces the behavior of the client application.
  • Test editing. After recording, you can edit the events that were recorded. You can replace recorded test values with variable test data or add dynamic data to the test.
  • Test validation. Before deploying the test, you can run the test manually as a single virtual user to make sure that the test runs smoothly and produces the expected results in a nominal environment with minimal server load. You might experience multiple test editing and validation cycles before your test runs as expected.
  • Workload emulation with schedules. When the test runs repeatedly as anticipated, you specify an execution schedule and user groups to emulate a workload that a large number of virtual users generates.
  • Schedule execution. You run the schedule, deploying test execution over virtual users that can be hosted on remote hosts. Each virtual user runs an instance of the test. Response time results are collected.
  • Evaluation of results. You evaluate the results produced by the tests through the various reports that are generated during execution. You can also design custom reports.