Skip to content

collectors NagiosPerfdataCollector

Rob Smith edited this page Sep 16, 2014 · 5 revisions

NagiosPerfdataCollector

The NagiosPerfdataCollector parses Nagios performance data in the PNP4Nagios/Graphios/Metricinga key-value format.

Dependencies

  • Nagios configured to periodically dump performance data files in PNP4Nagios format

Configuring Nagios/Icinga

If you're already using Graphios, you're already set up to send metrics through Metricinga, and you can skip to the next section! If not, read on.

Modifying the daemon configuration

The default performance data output format used by Nagios and Icinga can't be easily extended to contain new attributes, so we're going to replace it with one that prints key-value pairs instead. This will allow us to add in whatever kind of bookkeeping attributes we want! We need these to do things like override the display name of a service with a metric name more meaningful to Graphite.

We'll need to edit one of the following files:

  • For Nagios: /etc/nagios/nagios.cfg
  • For Icinga: /etc/icinga/icinga.cfg

Make sure that the following configuration keys are set to something like the values below:

process_performance_data=1
host_perfdata_file=/var/spool/nagios/host-perfdata
host_perfdata_file_mode=a
host_perfdata_file_processing_command=process-host-perfdata-file
host_perfdata_file_processing_interval=60
host_perfdata_file_template=DATATYPE::HOSTPERFDATA	TIMET::$TIMET$	HOSTNAME::$HOSTNAME$	HOSTPERFDATA::$HOSTPERFDATA$	HOSTCHECKCOMMAND::$HOSTCHECKCOMMAND$	HOSTSTATE::$HOSTSTATE$	HOSTSTATETYPE::$HOSTSTATETYPE$	GRAPHITEPREFIX::$_HOSTGRAPHITEPREFIX$	GRAPHITEPOSTFIX::$_HOSTGRAPHITEPOSTFIX$  # NOQA
service_perfdata_file=/var/spool/nagios/service-perfdata
service_perfdata_file_mode=a
service_perfdata_file_processing_command=process-service-perfdata-file
service_perfdata_file_processing_interval=60
service_perfdata_file_template=DATATYPE::SERVICEPERFDATA	TIMET::$TIMET$	HOSTNAME::$HOSTNAME$	SERVICEDESC::$SERVICEDESC$	SERVICEPERFDATA::$SERVICEPERFDATA$	SERVICECHECKCOMMAND::$SERVICECHECKCOMMAND$	HOSTSTATE::$HOSTSTATE$	HOSTSTATETYPE::$HOSTSTATETYPE$	SERVICESTATE::$SERVICESTATE$	SERVICESTATETYPE::$SERVICESTATETYPE$	GRAPHITEPREFIX::$_SERVICEGRAPHITEPREFIX$	GRAPHITEPOSTFIX::$_SERVICEGRAPHITEPOSTFIX$  # NOQA

Note that you most likely will wish to change $_SERVICEGRAPHITEPREFIX$, $_HOSTGRAPHITEPREFIX$, $_SERVICEGRAPHITEPOSTFIX$, and $_HOSTGRAPHITEPOSTFIX$

Configuring file rotation

Next, the rotation commands need to be configured so the performance data files are periodically moved into the Metrnagios spool directory. Depending on your system configuration, these commands may be located in /etc/nagios/objects/commands.d:

define command {
    command_name    process-host-perfdata-file
    command_line    /bin/mv /var/spool/nagios/host-perfdata /var/spool/diamond/host-perfdata.$TIMET$  # NOQA
}

define command {
    command_name    process-service-perfdata-file
    command_line    /bin/mv /var/spool/nagios/service-perfdata /var/spool/diamond/service-perfdata.$TIMET$  # NOQA
}

Options - Generic Options

Setting Default Description Type
byte_unit byte Default numeric output(s) str
enabled False Enable collecting these metrics str
measure_collector_time False Collect the collector run time in ms bool
metrics_blacklist None Regex to match metrics to block. Mutually exclusive with metrics_whitelist NoneType
metrics_whitelist None Regex to match metrics to transmit. Mutually exclusive with metrics_blacklist NoneType
perfdata_dir /var/spool/diamond/nagiosperfdata The directory containing Nagios perfdata files str

Example Output

__EXAMPLESHERE__

This file was generated from the python source

Please edit the source to make changes

Clone this wiki locally