The Diagnostics Collector has several settings that affect
the way the collector works.
The settings file uses the standard Java™ properties format. It is a text file with
one property=value pair on each line. Each supported
property controls the Diagnostics Collector in some way. Lines that
start with '#' are comments.
Parameters
- file.<any_string>=<pathname>
Any property with a name starting file. specifies
the path to a diagnostics file to collect. You can add any string
as a suffix to the property name, as a reminder of which file the
property refers to. You can use any number of file. properties,
so you can tell the Diagnostics Collector to collect a list of custom
diagnostic files for your environment. Using file. properties
does not alter or prevent the collection of all the standard diagnostic
files. Collection of standard diagnostic files always takes place.
Custom
debugging scripts or software can be used to produce extra output
files to help diagnose a problem. In this situation, the settings
file is used to identify the extra debug output files for the Diagnostics
Collector. The Diagnostics Collector collects the extra debug files
at the point when a problem occurs. Using the Diagnostics Collector
in this way means that debug files are collected immediately after
the problem event, increasing the chance of capturing relevant context
information.
- output.dir=<output_directory_path>
The Diagnostics Collector tries to write its output .zip file
to the output directory path that you specify. The path can be absolute
or relative to the working directory of the Java process. If the directory does not exist,
the Diagnostics Collector tries to create it. If the directory cannot
be created, or the directory is not writeable, the Diagnostics Collector
defaults to writing its output .zip file to the current working directory.
Note: On
Windows systems, Java properties
files use backslash as an escape character. To specify a backslash
as part of Windows® path
name, use a double backslash '\\' in the properties file.
- loglevel.file=<level>
- This setting controls the amount of information written to the
Diagnostics Collector log file. The default setting for this property
is config. Valid levels are:
- off
- No information reported.
- severe
- Errors are reported.
- warning
- Report warnings in addition to information reported by severe.
- info
- More detailed information in addition to that reported by warning.
- config
- Configuration information reported in addition to that reported
by info. This is the default reporting level.
- fine
- Tracing information reported in addition to that reported by config.
- finer
- Detailed tracing information reported in addition to that reported
by fine.
- finest
- Report even more tracing information in addition to that reported
by finer.
- all
- Report everything.
- loglevel.console=<level>
- Controls the amount of information written by the Diagnostics
Collector to stderr. Valid values for this property
are as described for loglevel.file.
The default setting for this property is warning.
- settings.id=<identifier>
- Allows you to set an identifier for the settings file. If you
set loglevel.file to fine or lower,
the settings.id is recorded in the Diagnostics
Collector log file as a way to check that your settings file is loaded
as expected.
- config.check.javacore={true|false}
- Set config.check.javacore=true to enable
a Java dump for the diagnostics
configuration check at virtual machine start-up. The check means that
the virtual machine start-up takes more time, but it enables the most
thorough level of diagnostics configuration checking.
- run.jextract=false
- Set this option to prevent the Diagnostics Collector running jextract on
detected System dumps.