Tool Mentor: Create a Performance Schedule Using IBM Rational Performance Tester
This tool mentor describes how to use IBM Rational Performance Tester to create a performance test schedule.
Tool: IBM Rational Performance Tester
Relationships
Main Description

A schedule allows you to order, group, and run tests. With schedules, you can run tests remotely, establish the order of tests (sequential, random, or weighted order), run tests at a specific rate, or specify the number of times a given test runs. You can also use schedules to run tests against specified builds.

IBM® Rational® Performance Tester provides a convenient wizard for creating schedules. To create a performance schedule, right-click your project and then click New > Performance Schedule.

In the Performance Schedule Wizard, enter the name of the schedule and click Finish. At this point, a new schedule with one user group appears. User groups allow you to create logical groupings by using various characteristics representing various types of users on your system. In order to achieve a more realistic distribution of users, you may want to create various groups to represent different categories, such as buyers, sellers, users checking on the status of an order, or browsers. The user group types will be determined by the characteristic of the users of your system.

In the Perforance Schedule Wizard, you can also easily add tests. With the schedule you just created open, right-click the user group and select Add > Test. In the Select Performance Test window, select the test you want to add to your schedule. You can also alternate start times for users in the User Load section of the schedule by selecting the check box Add a delay between starting each user.

In general, when you are building a schedule, you will want to achieve a random arrival delay with every transaction. You will also want to prevent authentication rejections by ramping up user numbers slowly, or by starting with random delays. You can, for example, start with 1 to 4 users per second and ramp-up, or start with 10 to 20 with random delays over several minutes. Finally, when you are building your schedule, make sure that the test runs long enough to get steady state results. This can be achieved by making the workload last a long time relative to ramp-up.

For more information on working with schedules using Rational Performance Tester, see the Tool Mentor: Edit Schedules with IBM Rational Performance Tester.