WebSphere Message Service Client for C/C++, Version 2.0.2 Operating Systems: AIX, Linux, Solaris, Windows

Repeatable failures

If you are dealing with a repeatable failure, it might be necessary for you to capture product trace over an extended period of time to allow the problem to be diagnosed.

If you need to provide a product trace, either enable trace as advised by the IBM Support Center representative or as described in FFDC and trace configuration for C/C++ applications.

It is important that the size of the trace file is large enough to capture the trace while the repeatable problem occurs. To set the size of the trace file, either use environment variable XMS_TRACE_FILE_SIZE or use the gxisc executable command as follows:

alter trace(enabled) tracesize(xxxx)

Refer to Table 1 for the descriptions of various environment variable settings for C/C++ trace.

After the failure that you are tracing has occurred, you must either copy the trace files or disable the trace using the following command:

gxisc trace(disabled)

This is because trace wraps, which means that leaving trace on would eventually cause the trace at the point of failure to be lost.

XMS product trace is written in compressed form to gain a performance advantage. You can format the trace files using the gxitrcfmt tool.

If you are experiencing problems where you do not have access to the information provided in the error block, you may want to enable the XMS_FFDC_EXCEPTIONS environment variable. This produces an FFDC record whenever the XMS API returns an error from a function. The FFDC record contains full details of the XMS error block to assist in debugging failures.

Refer to FFDC and trace configuration for C/C++ applications for more information on the gxisc and gxitrcfmt commands.


Concept topic

Terms of Use | Rate this page

Last updated: 24 May 2011

(C) Copyright IBM Corporation 2005, 2011. All Rights Reserved.