7.15. Performance Data

7.15.1. Introduction
7.15.2. Types of Performance Data
7.15.3. Plugin Performance Data
7.15.4. Processing Performance Data
7.15.5. Processing Performance Data Using Commands
7.15.6. Writing Performance Data To Files
7.15.7. Performance Data Format

7.15.1. Introduction

Icinga is designed to allow plugins to return optional performance data in addition to normal status data, as well as allow you to pass that performance data to external applications for processing. A description of the different types of performance data, as well as information on how to go about processing that data is described below...

7.15.2. Types of Performance Data

There are two basic categories of performance data that can be obtained from Icinga:

  1. Check performance data

  2. Plugin performance data

Check performance data is internal data that relates to the actual execution of a host or service check. This might include things like service check latency (i.e. how "late" was the service check from its scheduled execution time) and the number of seconds a host or service check took to execute. This type of performance data is available for all checks that are performed. The $HOSTEXECUTIONTIME$ and $SERVICEEXECUTIONTIME$ macros can be used to determine the number of seconds a host or service check was running and the $HOSTLATENCY$ and $SERVICELATENCY$ macros can be used to determine how "late" a regularly-scheduled host or service check was.

Plugin performance data is external data specific to the plugin used to perform the host or service check. Plugin-specific data can include things like percent packet loss, free disk space, processor load, number of current users, etc. - basically any type of metric that the plugin is measuring when it executes. Plugin-specific performance data is optional and may not be supported by all plugins. Plugin-specific performance data (if available) can be obtained by using the $HOSTPERFDATA$ and $SERVICEPERFDATA$ macros. Read on for more information on how plugins can return performance data to Icinga for inclusion in the $HOSTPERFDATA$ and $SERVICEPERFDATA$ macros.

7.15.3. Plugin Performance Data

At a minimum, Icinga plugins must return a single line of human-readable text that indicates the status of some type of measurable data. For example, the check_ping plugin might return a line of text like the following:

 PING ok - Packet loss = 0%, RTA = 0.80 ms 

With this simple type of output, the entire line of text is available in the $HOSTOUTPUT$ or $SERVICEOUTPUT$ macros (depending on whether this plugin was used as a host check or service check).

Plugins can return optional performance data in their output by sending the normal, human-readable text string that they usually would, followed by a pipe character (|), and then a string containing one or more performance data metrics. Let's take the check_ping plugin as an example and assume that it has been enhanced to return percent packet loss and average round trip time as performance data metrics. Sample output from the plugin might look like this:

 PING ok - Packet loss = 0%, RTA = 0.80 ms | percent_packet_loss=0 rta=0.80

When Icinga sees this plugin output format it will split the output into two parts:

  1. Everything before the pipe character is considered to be the "normal" plugin output and will be stored in either the $HOSTOUTPUT$ or $SERVICEOUTPUT$ macro

  2. Everything after the pipe character is considered to be the plugin-specific performance data and will be stored in the $HOSTPERFDATA$ or $SERVICEPERFDATA$ macro

In the example above, the $HOSTOUTPUT$ or $SERVICEOUTPUT$ macro would contain "PING ok - Packet loss = 0%, RTA = 0.80 ms" (without quotes) and the $HOSTPERFDATA$ or $SERVICEPERFDATA$ macro would contain "percent_packet_loss=0 rta=0.80" (without quotes).

Multiple lines of performance data (as well as normal text output) can be obtained from plugins, as described in the plugin API documentation.

[Note] Note

The Icinga daemon doesn't directly process plugin performance data, so it doesn't really care what the performance data looks like. The developer guidelines contain several rules (see performance data format below for details). However, if you are using an external addon to process the performance data (i.e. PerfParse), the addon may be expecting that the plugin returns performance data in a specific format. Check the documentation that comes with the addon for more information.

7.15.4. Processing Performance Data

If you want to process the performance data that is available from Icinga and the plugins, you'll need to do the following:

  1. Enable the process_performance_data option.

  2. Configure Icinga so that performance data is either written to files and/or processed by executing commands.

Read on for information on how to process performance data by writing to files or executing commands.

7.15.5. Processing Performance Data Using Commands

The most flexible way to process performance data is by having Icinga execute commands (that you specify) to process or redirect the data for later processing by external applications. The commands that Icinga executes to process host and service performance data are determined by the host_perfdata_command and service_perfdata_command options, respectively.

An example command definition that redirects service check performance data to a text file for later processing by another application is shown below:

 define command{
        command_name    store-service-perfdata
[Tip] Tip

This method, while flexible, comes with a relatively high CPU overhead. If you're processing performance data for a large number of hosts and services, you'll probably want Icinga to write performance data to files instead. This method is described in the next section.

7.15.6. Writing Performance Data To Files

You can have Icinga write all host and service performance data directly to text files using the host_perfdata_file and service_perfdata_file options. The format in which host and service performance data is written to those files is determined by the host_perfdata_file_template and service_perfdata_file_template options.

An example file format template for service performance data might look like this:

[Note] Note

The above is just one example of how to define a template. This definition will NOT work for PNP!

By default, the text files will be opened in "append" mode. If you need to change the modes to "write" or "non-blocking read/write" (useful when writing to pipes), you can use the host_perfdata_file_mode and service_perfdata_file_mode options.

Additionally, you can have Icinga periodically execute commands to periocially process the performance data files (e.g. rotate them) using the host_perfdata_file_processing_command and service_perfdata_file_processing_command options. The interval at which these commands are executed are governed by the host_perfdata_file_processing_interval and service_perfdata_file_processing_interval options, respectively.

7.15.7. Performance Data Format

The Monitoring Plugins Developer Guidelines (https://www.monitoring-plugins.org/doc/guidelines.html) contain in section 2.6 the specification and an explanation of the format of performance data. It is repeated below for your convenience:

----- 8< -----

2.6.Performance data

Nagios 3 and newer will concatenate the parts following a "|" in a) the first line output by the plugin, and b) in the second to last line, into a string it passes to whatever performance data processing it has configured. (Note that it currently does not insert additional whitespace between both, so the plugin needs to provide some to prevent the last pair of a) and the first of b) getting run together.) Please refer to the Nagios documentation for information on how to configure such processing. However, it is the responsibility of the plugin writer to ensure the performance data is in a "Nagios Plugins" format. This is the expected format:



  1. space separated list of label/value pairs

  2. label can contain any characters

  3. the single quotes for the label are optional. Required if spaces, = or ' are in the label

  4. label length is arbitrary, but ideally the first 19 characters are unique (due to a limitation in RRD). Be aware of a limitation in the amount of data that NRPE returns to Nagios

  5. to specify a quote character, use two single quotes

  6. warn, crit, min, and/or max, respectively, may be null (for example, if the threshold is not defined or min and max do not apply). Trailing unfilled semicolons can be dropped

  7. min and max are not required if UOM=%

  8. value, min and max in class [-0-9.]. Must all be the same UOM

  9. warn and crit are in the range format (see Section 2.5). Must be the same UOM

  10. UOM (unit of measurement) is one of:

    • no unit specified - assume a number (int or float) of things (eg, users, processes, load averages)

    • s - seconds (also us, ms)

    • % - percentage

    • B - bytes (also KB, MB, TB, GB?)

    • c - a continous counter (such as bytes transmitted on an interface)

It is up to third party programs to convert the Nagios plugins performance data into graphs.

----- 8< -----

Every now and then people are curious if their performance data is valid so we'll give some examples showing the data following the pipe symbol:

 loss=0 rta=0.80ms         # valid (rule 1,10)
 loss=0, rta=0.80ms        # invalid (rule 1, no comma allowed)
 loss=0 rta=0,80ms         # invalid (rule 8, no comma allowed) *1
 packet loss=0 rta=0.80    # invalid (rule 3, single quotes needed)
 'packet loss'=0 rta=0.80  # valid version of line above
 'john''s disk'=83%        # valid (rule 5,10)
 'disk usage'=78%;80;90;;; # invalid (trailing semicolon after max value)
 'disk usage'=78%;80;90    # valid (rule 6)
 'data packets'=11345234c  # valid (rule 6,10)
 drum=153482pages          # invalid (rule 10, wrong UOM) *2
 temperature=23;;;20;30    # valid (rule 6)
[Note] Note

*1: Having a decimal comma instead of a decimal point may be a result of using your local language settings. It might be fixed internally by your graphing tool but don't rely on it.

*2: Having "pages" or other non-standard UOMs (e.g. °C/°F) will probably result in errors or discarded performance data when using addon like PNP4Nagios or other tools which try to graph your performance data.