Configuring settings after recording HTTP scripts for response time reporting

Note: For more information, consult the Rational® Performance Tester product online help.

The following sections describe several best practice settings that you should configure after you record Rational Performance Tester (RPT) HTTP scripts for response time reporting in the Tivoli Enterprise Portal.

Enable response time breakdown

Use the following procedure to enable response time breakdown, which generates response time information for page elements and subtransactions (note that the plugin typically does this for you automatically when the script is uploaded):
  1. After recording the RPT HTTP script, open the script in the Rational Performance Tester workbench Test Perspective.
  2. Select the parent node.
  3. Select the Enable response time breakdown checkbox.

These steps are illustrated in the following example:

Example of enabling response time breakdown in RPT Test Perspective

Scale back browser delays

Scaling the browser delays to zero can have a significant impact on the reported client time. When you record a script, RPT saves a browser delay factor when RPT determines that the browser is waiting for some action to complete and that action is not related to the current page element. In some instances these browser delays can account for a few milliseconds up to several seconds. If you believe the client response time is unusually high, you should inspect the browser delays recorded in the script.

Use the following procedure to view and adjust the delay recorded for a given page element:
  1. After recording the RPT HTTP script, open the script in the Rational Performance Tester workbench Test Perspective.
  2. Select the parent node.
  3. Select the HTTP Options tab, as shown in the following example:
    Example of selecting the HTTP Options tab in the RPT Test Perspective.
  4. Scroll through the page elements one at a time. Scroll down to inspect the Delay field. In this example, no delays were recorded. If you set this field to a value and save changes, this value is permanently recorded with the script.
    Example of checking the Delay field in the RPT Test Perspective.
  5. Move the slider all the way to the left. This step globally scales the browser delays to zero in all of the page elements.
    Example of adjusting the slider to the shortest delay in the RPT Test Perspective.
  6. When you select a delay scale factor, the following message is displayed. Click OK.
    Example of the message that is displayed when you select a delay scale factor in the RPT Test Perspective.

Note that these browser delays can be scaled by percentages. For example, if the recorded delay is 400 milliseconds, scaling back the delay to 10% will cause the playback to insert a 40 millisecond delay for the current page element. The original value and the scaling factor are preserved. Consequently, you can later restore the original delay factor by clicking the slider Reset button. Or, you can change the scaling factor by simply moving the slider.

Ignore embedded object response times

Use the following procedure to ignore any response times associated with embedded objects, such as images:
  1. After recording the RPT HTTP script, open the script in the Rational Performance Tester workbench Test Perspective.
  2. Select the parent node.
  3. Select the HTTP Options tab.
  4. Adjust the Playback speed by moving the slider all the way to the left.
  5. Select the Enable response time breakdown checkbox
  6. Click Modify.
    Example of performing the previous steps in the RPT Test Perspective.
  7. The Test Editor displays. Configure the script to ignore embedded object response times (such as images). The following example shows the options that are available for selection. For example, enable All Secondary.
    Example of ignoring embedded object response times in the RPT Test Perspective.

Set retries to 0, increase timeout value in the script profile

If you start observing script timeout errors in the Tivoli Enterprise Portal (TEP) workspaces, increase the Timeout Period in the script profile, as shown (1) in the following example. This best practice ensures that the RPT engine times out the script instead of ITCAM for Transactions. The Timeout Period is the ITCAM for Transactions timeout period, not the timeout period defined in the RPT script.

Example of configuring script properties

In addition to increasing the timeout period, set the Number of Retries (2) to zero. If the RPT engine detects a script error, the RPT engine reports the error immediately and stops the script run. This best practice ensures that you are notified immediately when script errors occur without inflating the client time and overall response time.