InfoCenter Home >
6: Administer applications >
6.6: Tools and resources quick reference >
6.6.19: Administering the product messages, logs, and traces (overview) >
6.6.19.1: Administering the product messages, logs, and traces

6.6.19.1: Administering the product messages, logs, and traces

The Application Server product provides various levels of information, from high-level messages in the console to more detailed server execution logs and comprehensive tracing.

General approach to tracing a problem

You should not need to trace IBM WebSphere Application Server unless directed to do so by IBM support personnel. In case you would like to perform your own tracing, here is a procedure:

  1. Determine the activity during which the problem usually occurs.

  2. Configure an application server for a dump, but do not trigger the dump yet.
  3. Perform the activity during which the problem usually occurs.

  4. After the problem occurs, proceed with the dump.

  5. Check the dump file, which will contain only the tracing performed between the time you specified the classes to trace and the moment when you dumped the trace data to the file.

  6. Optionally, based on the dump contents, speculate which server classes or components are involved in the problem. Repeat the procedure, this time narrowing the scope of the trace to just those components, isolating the problem area.

Dump the server trace to a file

Sometimes the administrator might find it useful to dump the contents of the ring buffer that holds trace data for an administrative or application server.

  1. Display the Topology tree.

  2. To access the Trace Administration dialog box, right-click a running node or application server in the Topology tree and click Trace.
  3. In the Trace Administration dialog box, specify the classes to trace. Verify the dump file name.

  4. Now or later, click the Dump button.

    To isolate the activity you are trying to trace, you might want to delay clicking the Dump button until after the server performs the activity. See the general trace approach section above for more details about this strategy.

Trace an administrative server from startup

To trace an administrative server from the time that it starts, modify its properties related to tracing:

  1. Open the admin.config file in the bin directory of the WebSphere Application Server installation root.

  2. Set the traceString property to specify the server classes to trace. For valid syntax, see the trace properties help.

  3. Set the traceOutput property to specify the file to which to send the trace data. For valid syntax, see the server trace properties help.

  4. Save the file.

  5. Start the server (or stop it and start it again) for the new trace parameters to take effect.

Trace an application server from startup

To trace an application server from the time that it starts, modify its properties related to tracing:

  1. Display the Topology tree.

  2. In the tree, locate and click the application server to display its properties on the right side of the console.

  3. Set the trace-related properties. Consult the server trace properties help for syntax.

  4. Save the property changes.

  5. Start the server (or stop it and start it again) for the new trace parameters to take effect.

Go to previous article: Server trace properties Go to next article: Administering messages with the Java administrative console

 

 
Go to previous article: Server trace properties Go to next article: Administering messages with the Java administrative console