Tracing the plug-in

Tracing is enabled in the plug-in by setting LogLevel to Trace, as specified in 5.6.1, Logging.

A good method for starting a new trace is to change the LogLevel to Trace and change the name of the file to a new value, for example traceplugin.log. This way, you know where your trace starts and you have an easily managed trace file.

Once you are finished, you can return the plug-in configuration file to its original settings and do not need to stop the Web server or the appserver for any of this to work.

The trace itself is quite straightforward to follow. If you are tracing requests, you will see a request handled as shown in Figure 5-7.

Tracing will provide you with information about: The virtual hosts the plug-in is using. You can view how the virtual host is matched to the request. One of the most common problems is specifying the same alias in multiple virtual hosts and so not giving the plug-in a unique route to follow. The URIs the plug-in is searching. If you are unsure as to why a request is not reaching your application, use a trace to see how the plug-in is matching your browser request. Workload management selection process. Tracing is useful if you wish to validate how your cluster members are being selected and watch failover mechanisms operate. Observing the transport mechanism being used and connections or failures in connection. Session management.

  Prev | Home | Next

 

WebSphere is a trademark of the IBM Corporation in the United States, other countries, or both.

 

IBM is a trademark of the IBM Corporation in the United States, other countries, or both.