Trace File Location

Trace files are collected per execution group and currently persisted to the working directory of the broker by default.

This can be changed by setting the following entry in the VALUE table:

Table 1. Trace location entry in the VALUE table
Category	Key	Value
CORE	PROFILING_FOLDER_NAME	directory name

The following table shows the naming convention for trace files. The <n> in the file naming pattern indicates a numeric, 'rolling' value.

Table 2. Trace file locations
Trace type	File names
Basic statistics	<broker name>_<execution group name>_basic.log.<n>
Full	<broker name>_<execution group name>_trace.log.<n>
Service	<broker name>_<execution group name>_srvtrc.log.<n>

Basic statistics trace files are written a short time after the monitoring module detects inactivity (approximately 45 seconds). This allows the data to be collected in memory, which minimizes overhead, and then processed and written when the test is complete.

The full trace is written using an internal buffer that is flushed when the buffer is full. This means that the full trace file is written near real time in a busy system. The file is flushed at the same time the basic statistics file is written to ensure that all data is written.

The service trace is written using an internal buffer that is flushed when the buffer is full. This means that the service trace file is written near real time in a busy system.