This section describes:
Overview of problem diagnosis
WebSphere Application Server for z/OS uses component trace (CTRACE) to capture and display trace data in trace data sets. WebSphere Application Server for z/OS identifies itself to CTRACE with a dynamic component name determined by the short cell name. CTRACE allows you to:
WebSphere Application Server for z/OS also has an error log stream that records error information when WebSphere Application Server for z/OS detects an unexpected condition or failure within its own code, such as:
Use the error log stream in conjunction with other facilities available to capture error or status information, such as an activity log, trace data, system logrec, and job log.
The WebSphere Application Server for z/OS error log stream is a system logger application. Because the error log stream uses the system logger, you can:
Note: There is a significant performance penalty when using DASD-only error logging.
WebSphere Application Server for z/OS provides a REXX EXEC (BBORBLOG) that allows you to browse the error log stream. By default, the EXEC formats the error records to fit a 3270 display.
This topic describes the error
log stream and how to set it up. Information about using the error log stream
to diagnose problems is in the Troubleshooting section of the WebSphere Application
Server information center. General information and guidance about the system
logger is in z/OS MVS Setting Up a Sysplex.
For details about problem diagnosis, see the Troubleshooting section of the WebSphere Application Server information center.
Planning for Component Trace
To use CTRACE, you:
Recommendation for dumps
Plan as you would normally for system dumps. Due to the size of WebSphere Application Server for z/OS address spaces, you may need to re-size your system dump data sets and use dynamic dump data sets.