22. Configuration Overview
NXLog uses
Apache style
configuration files. The configuration file is loaded from its default
location, or it can be explicitly specified with the -c
command line
argument.
The configuration file is made up of blocks and directives. Blocks are
similar to XML tags containing multiple directives. Directive names
are case insensitive but arguments are case sensitive in some cases. A
directive and its argument must be specified on the same line. Values
spanning multiple lines must have the newline escaped with the
backslash (\
). A typical case for this is the
Exec directive. Blank lines and lines starting
with the hash mark (#
) are ignored. Configuration directives
referring to a file or a path can be quoted with double quotes ("
)
or single quotes ('
). This applies to both global and module
directives.
The configuration file can be logically divided into three parts: global parameters, module instances, and route instances.
Here are the three sections of a configuration file. This example contains two global directives, Input and Output blocks, and a Route block.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# Global section
User nxlog
Group nxlog
# Modules section
<Input in>
Module im_null
</Input>
<Output out>
Module om_null
</Output>
# Route section
<Route r>
Path in => out
</Route>
22.1. Global Directives
The global section contains directives that control the overall behavior of NXLog.
The LogFile directive sets a destination file for NXLog internal logs. If this directive is unset, the log file is disabled and internal NXLog logs are not written to file (unless configured via the im_internal module). See also Rotating the Internal Log File.
With the User and Group directives set, NXLog will drop root privileges after starting and run under the specified user and group. These directives are ignored if running on the Windows platform.
After starting, NXLog will change its working directory to the directory specified by the SpoolDir directive. Non-absolute paths in the configuration will be relative to this directory.
See the Reference Manual for a complete list of available global directives.
22.2. Modules
NXLog will only load modules which are specified in the
configuration file and used in an active route.
A module instance is specified according to its corresponding module type
(Extension
, Input
, Processor
, or Output
). Each module instance must
have a unique name and a Module directive. The
following is a skeleton configuration block for an input module.
1
2
3
4
<Input instancename>
Module im_module
...
</Input>
For more details about module instance names, see Configuration in the Reference Manual.
22.3. Routes
Routes define the flow and processing order of the log messages. Each route instance must have a unique name and a Path.
This Route instance, named example
, takes logs from Input module
instances named in1
and in2
, processes the logs with the proc
Processor module instance, and sends the resulting logs to both Output
module instances out1
and out2
. These named module instances must
be defined elsewhere in the configuration file.
1
2
3
<Route example>
Path in1, in2 => proc => out1, out2
</Route>
For more details about route instance names, see Configuration in the Reference Manual.
If no Route block is specified in the configuration, NXLog will automatically generate a route, with all the Input and Output instances specified in a single path.
NXLog can use a configuration with no Route block, such as the following.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
<Input in1>
Module im_null
</Input>
<Input in2>
Module im_null
</Input>
<Output out1>
Module om_null
</Output>
<Output out2>
Module om_null
</Output>
The following Route block will be generated automatically.
1
2
3
<Route r>
Path in1, in2 => out1, out2
</Route>
22.4. Constant and Macro Definitions
A define is useful if there are many instances in the configuration where the same value must be used. Typically, defines are used for directories and hostnames. In such cases the value can be configured with a single definition. In addition to constants, other strings like code snippets or parser rules can be defined in this way.
An NXLog define works in a similar way to the C language, where the pre-processor substitutes the value in places where the macro is used. The NXLog configuration parser replaces all occurrences of the defined name with its value, and then after this substitution the configuration check occurs.
The define directive can be used for statements as
shown above, but multiple statements should be specified using a code
block, with curly braces ({}
), to result in the expected
behavior.
The following example shows an incorrect use of the define directive. After substitution, the drop() procedure will always be executed; only the warning message will be logged conditionally.
1
2
3
4
5
6
7
define ACTION log_warning("dropping message"); drop();
<Input in>
Module im_file
File '/var/log/messages'
Exec if $raw_event =~ /dropme/ %ACTION%
</Input>
To avoid this problem, the action should be defined using a code block.
1
define ACTION { log_warning("dropping message"); drop(); }
22.5. Environment Variables
The envvar directive works like
define except that the value is retrieved
from the environment. This makes it possible to reference the
environment variable as if it was a define
. This directive is only
available in NXLog Enterprise Edition.
This is similar to the previous example using a define, but here the value is fetched from the environment.
22.6. File Inclusion
NXLog provides several features for including configuration directives and blocks from separate files or from executables.
Note
|
The SpoolDir directive does not take effect until after the configuration has been parsed, so relative paths specified with these directives are relative to the working directory where NXLog was started from. Generally, it is recommended to use absolute paths. If desired, define directives can be used to simulate relative paths (see Using Defines to Include a Configuration File). |
With the include directive it is possible to specify a file or set of files to be included in the current configuration file.
This example includes the contents of the /opt/nxlog/etc/syslog.conf
file in
the current configuration.
1
include /opt/nxlog/etc/syslog.conf
In this example for Windows, two define directives are
used to include an eventlog.conf
configuration file.
1
2
3
define ROOT C:\Program Files (x86)\nxlog
define CONFDIR %ROOT%\conf
include %CONFDIR%\eventlog.conf
The include directive also supports wildcarded
filenames. A set of files in a directory can be included without the need to
explicitly list each one. This could be used to specify a drop-in directory for
OS-specific configuration snippets (like windows2003.conf
) or
application-specific snippets (such as syslog.conf
).
This example includes all .conf
files in /opt/nxlog/etc/nxlog.d
.
1
include /opt/nxlog/etc/nxlog.d/*.conf
Because the backslash (\
) is used as a path separator on Windows, an
additional backslash (\
) must be used to prevent the \*
from being
treated as a literal asterisk (*
). The same wildcard rules apply as for the
im_file File directive.
1
include C:\Program Files\nxlog\conf\nxlog.d\\*.conf
With the include_stdout directive, an external command can be used to provide configuration content. There are many ways this could be used, including fetching, decrypting, and validating a signed configuration from a remote host, or generating configuration content dynamically.
Here, a separate script is responsible for fetching the NXLog configuration.
1
include_stdout /opt/nxlog/etc/fetch_conf.sh