First steps with nagiosplugin

This tutorial will guide you through all important steps of writing a check with the nagiosplugin class library. Read this to get started.

Key concepts

nagiosplugin has a fine-grained class model with clear separation of concerns. This allows plugin writers to focus on one particular tasks at a time while writing plugins. Nagios/Icinga plugins need to perform three steps: data acquisition, evaluation, and presentation. Each step has an associated class (Resource, Context, Summary) and information between tasks is passed with structured value objects (Metric, Result).

Classes overview

Here is a diagram with the most important classes and their relationships:

             +----------+                \
             | Resource |                 |
             +----------+                 |
       _____/      |     \_____           | Acquisition
      v            v           v          |
+---------+   +---------+   +---------+   |
| Metric  |...| Metric  |...| Metric  |  <
+---------+   +---------+   +---------+   |
     |             |             |        |
     v             v             v        |
+---------+   +---------+   +---------+   |
| Context |...| Context |...| Context |   | Evaluation
+---------+   +---------+   +---------+   |
     |             |             |        |
     v             v             v        |
+---------+   +---------+   +---------+   |
| Result  |...| Result  |...| Result  |  <
+---------+   +---------+   +---------+   |
           \___    |    ___/              |
               v   v   v                  | Presentation
              +---------+                 |
              | Summary |                 |
              +---------+                /

A model of the thing being monitored. It should usually have the same name as the whole plugin. Generates one or more metrics.

Example: system load


A single measured data point. A metric consists of a name, a value, a unit, and optional minimum and maximum bounds. Most metrics are scalar (the value can be represented as single number).

Example: load1=0.75


Additional information to evaluate a metric. A context has usually a warning and critical range which allows to determine if a given metric is OK or not. Contexts also include information on how to present a metric in a human-readable way.

Example: warning=0.5, critical=1.0


Product of a metric and a context. A result consists of a state (“ok”, “warning”, “critical”, “unknown”), some explanatory text, and references to the objects that it was generated from.

Example: WARNING - load1 is 0.75


Condenses all results in a single status line. The status line is the plugin’s most important output: it appears in mails, text messages, pager alerts etc.

Example: LOAD WARNING - load1 is 0.75 (greater than 0.5)

The following tutorials which will guide you through the most important features of nagiosplugin.


Study the source code in the nagiosplugin/examples directory for complete examples.