Help

To start a test run in JBlitz Professional, simply click on the green 'arrow' toolbar icon, select the main menu option Run->Start or press the F8 key.

You should see a moving indicator in the top right corner and an animated 'cogs' icon to the left to show that JBlitz is querying the web pages. The background color for the test cases shown in the 'Main Setup' tab should darken. You should also see the green arrow and most other controls on the panel become disabled. This disabling is normal. It makes things clearer by indicating what can and cannot be done at any given stage.

When JBlitz is running a test run, all the virtual users configured for all the test cases are started and will begin independently issuing request messages and downloading responses. Whilst the run is ongoing, you are not able to modify the test configuration settings (the test run takes a 'snapshot' of the settings when the run is started). 

The following feedback information is viewable during your test run:

• Test case and summary feedback: shown on the main screen on the 'Main Setup' tab. 
• Error indicators: shown against the icon for any web page that has encountered an error.
• Virtual user activity: visual feedback on how each virtual user is progressing through its test case.
• Log Console: simple window that captures log messages at a configurable level of detail.

During a test run, a full suite of information is gathered together and captured to form the test results. Various visual tools are available for analyzing your results and these are discussed further in the next chapter.

You can trial run any single test case at any time by selecting the 'Dry Run' option from the right mouse popup menu for the test case in question. This option will run all the configured virtual users through the test case using the settings currently specified, but will only perform a single iteration (once through the download sequence). Once the cycle is finished, you can inspect results, errors etc in the usual way and modify the test case accordingly.

You can debug any given single test case or all test cases together using the JBlitz debugger. This gives you valuable inside information on what is occurring at each step in the request / response cycle. Select the third button in the right hand gutter. Extensive information on how to debug your test cases using the inbuilt debugger is provided.

There are numerous ways in which your test run may end. The most simple way is simply to press the red 'stop' icon to the right of the green 'start' icon. This allows you to halt a test run manually at any time.

The 'Run Control' tab of the main screen allows you to configure a time limit, hit limit or error limit for your test runs. Alternatively, you can specify in the 'Detection Action' of the 'Error Handling' tab that JBlitz should stop the current test run when an error is encountered (Stop all virtual users).

It is possible to stop virtual users from the Virtual User Activity windows. If, during a test run, all virtual users are stopped in this way, the test run will end.

Customization modules have the power to end a test run at any time if they provide a custom download sequence. More information about how to program a customization module to achieve this is given here.

Note: the evaluation version of JBlitz has an inbuilt time limit for any test run of 10 minutes.

The feedback area to the right of the test cases can be cycled through different feedbacks by clicking on the small green 'cycle' button on the subordinate toolbar in the main 'Main Setup' tab or by using the right mouse menu on a test case. Current feedbacks defined comprise the download speed (in kbits per second), total number of hits and errors for that test case, the state of the virtual users (running / stopped) and the current operation being performed.

To the right of each test case, an area shows text output giving informational feedback. If a test run is ongoing, the text will update as new hits are made or errors encountered. To the bottom left corner of the screen is a summary feedback display.

Test case feedback comprises:

• Current task. For example bytes being downloaded or web page hit event. A hit is any web page downloaded without error.
• Number of hits and errors for that web page.
• Download speed for that web page in kbits / second.
• Virtual user count and started / stopped state.

Summary feedback comprises:

• Total number of hits and errors.
• Total amount of data downloaded for all web pages so far this session.
• Total elapsed time this session.
• Average download speed for all web pages in kbits / second.

The virtual user activity window shows the current state of each virtual user for a given test case. Graphical feedback for each virtual user is displayed comprising a user icon with a sequential number, a connection state icon, a download bar and a traffic light icon. NOTE that, due to the intensive nature of the screen updating occurring, that, for high download speeds, test runs within JBlitz will be slowed somewhat when the window is showing.

The window shows schematically the web server, and to the right, each virtual user as they move through their connection / download / disconnection cycle. The green lights shows what page is currently being downloaded and the download bar just below shows the progress as the response for each page is received. Tooltips show the number of hits and errors. To the right, a traffic light shows whether the virtual user is still running or has stopped. An exclamation mark icon is shown to the right of the traffic light if the virtual user encountered an error, and, again, the tooltip will detail this error.

You can achieve fine grained virtual user control by using the right mouse button on any given virtual user to bring up the popup menu. Whilst JBlitz is testing, this menu allows you to stop or, if already stopped, restart that virtual user. You can do this for each of the virtual users to dynamically alter the simulated loading.

Click the 'Virtual User Activity' icon on the subordinate toolbar in the 'Main Setup' tab, or click the right mouse button on any of the test cases and select 'Virtual User Activity...'. A window is shown displaying the virtual user activity for the selected test case and for the current run or most recent run.

In addition to the logging support provided as part of the test configuration (in the main screen, 'Logs' tab), JBlitz includes a Log Console where you can quickly view the request / response conversation on screen either during or after a test run. Various options are available:

• Output filter - you can optionally filter on a given single test case.
• Errors only - you can choose to log only error related messages.
• Log level - specify the desired level as one of terse, normal, detailed, verbose.
• Auto scroll - make JBlitz keep scrolling to the bottom of the log as messages are added so the latest messages are viewable.

The Log Console is designed to be a quick and simple log destination. It is limited to ten thousand lines of text. After that point, the text at the start of the log is removed as text is added to the end. To enable logging, make sure the 'Logging enabled' check-box in the top left corner is selected.

The log can be cleared by clicking on the button toward the bottom right hand side of the screen. You can do a simple word search by typing into the search field at the bottom left of the screen and then clicking on the 'binoculars' icon. You can also cut and past text from the log to edit it in a more fully featured editor.

Log data shown in the Log Console is lost when the application exits.

Click the 'Log Console' icon on the main application toolbar or use main menu Tools->Log Console... or hitting Ctrl+Shift+L. The Log Console window is shown. A single Log Console window is maintained for the whole application.