ethMon

ethMon is an Icinga2/Nagios compatible Python plugin for network traffic monitoring on a specific interface. Currently the carbon/graphite data is always in bytes.

Graph drawn from a collected data

.. image:: images/plugin_graph.png

In order to calculate the speed the plugin keeps the data from previous runs in ~/.ethMonCache.db

Installation

• Download ethMon to the Icinga/Nagios plugin directory (/usr/lib/nagios/plugins/ on Debian 10)
  • Link_
• Ensure that the executable bit of ethMon is set

Configuration

• In order to use the plugin with Icinga the following configuration is needed:

  • in /usr/share/icinga2/include/command-plugins.conf add the following

    .. code-block:: console

    template CheckCommand "ethMon-common" {
        command = [ PluginDir + "/ethMon" ]
        arguments = {
                "-i" = {
                    value = "$interface$"
                    description = "Interface which will be monitored"
                }
                "-w" = {
                    value = "$speed_warn$"
                    description = "Warning threshold"
                }
                "-c" = {
                    value = "$speed_crit$"
                    description = "Critical threshold"
                }
                "-s" = {
                    value = "$speed_scale$"
                    description = "Speed scaling"
                }
                "--interval" = {
                    value = "$chk_interval$"
                    description = "Interval between the checks (in seconds)"
                }
        }
        vars.speed_warn = 60
        vars.speed_crit = 100
        vars.speed_scale = "Mb"
        vars.chk_interval = 30
    }
    object CheckCommand "ethMon" {
        import "ethMon-common"
    }

    ..

  • then per host (or zone if the interface name match everywhere):

    .. code-block:: console

    apply Service "Traffic-enp3s5" {
        check_command = "ethMon"
        check_interval = 30
        vars.interface = "enp3s5"
    command_endpoint = host.vars.client_endpoint // Check executed on client node
    assign where host.vars.client_endpoint == "TheMonitoredHost" // The service is applied only to this node
    }

    ..

  • NOTE: The same value must be used for check_interval & vars.chk_interval

Support for threshold ranges

• Icinga threshold ranges documentation_

The plugin has partial support for the ranges described in the Icinga documentation. If used the ranges must be specified for both (warning & critical) parameters. Implemented are the following specifications:

• min:max - The value must be outside the range in order the alert to be triggered
• @min:max - The value must be inside the range in order the alert to be triggered
• min: - The value must be below min or above 10*10^256 in order the alert to be triggered
• @min: - The value must be between min and 10*10^256 in order the alert to be triggered
• :max - The value must be below 0 and max in order the alert to be triggered
• @:max - The value must be between 0 and max in order the alert to be triggered

Motivation

The plugin has been created as check_eth from Nagios Exchange refused to work in my current setup.