Forge Home


Configuration and management of InfluxData's Telegraf metrics collection agent


5,006 latest version

4.3 quality score

We run a couple of automated
scans to help you access a
module's quality. Each module is
given a score based on how well
the author has formatted their
code and documentation and
modules are also checked for
malware using VirusTotal.

Please note, the information below
is for guidance only and neither of
these methods should be considered
an endorsement by Puppet.

Version information

  • 2.0.1 (deleted)
  • 1.5.1 (latest)
released Jun 20th 2018
This version is compatible with:
  • Puppet Enterprise 2017.2.x, 2017.1.x, 2016.5.x, 2016.4.x
  • Puppet > = 3.0.0 < 5.0.0
  • , , , , ,

Start using this module

  • r10k or Code Manager
  • Bolt
  • Manual installation
  • Direct download

Add this module to your Puppetfile:

mod 'ouroboros8-telegraf', '1.5.1'
Learn more about managing modules with a Puppetfile

Add this module to your Bolt project:

bolt module add ouroboros8-telegraf
Learn more about using this module with an existing project

Manually install this module globally with Puppet module tool:

puppet module install ouroboros8-telegraf --version 1.5.1

Direct download is not typically how you would use a Puppet module to manage your infrastructure, but you may want to download the module in order to inspect the code.



ouroboros8/telegraf — version 1.5.1 Jun 20th 2018


Build Status

Table of Contents

  1. Overview
  2. Setup
  3. Usage
  4. Limitations
  5. Development


A reasonably simple yet flexible Puppet module to manage configuration of InfluxData's Telegraf metrics collection agent.


There's a couple of fairly standard dependencies for this module, as follows:

NB: On some apt-based distributions you'll need to ensure you have support for TLS-enabled repos in place. This can be achieved by installing the apt-transport-https package.

In addition, for Windows, the following dependencies must be met:


Telegraf's configuration is split into four main sections - global tags, options specific to the agent, input plugins, and output plugins. The documentation for these sections is here, and this module aims to be flexible enough to handle configuration of any of these stanzas.

To get started, Telegraf can be installed with a very basic configuration by just including the class:

include ::telegraf

However, to customise your configuration you'll want to do something like the following:

class { '::telegraf':
    hostname => $::hostname,
    outputs  => {
        'influxdb' => {
            'urls'     => [ "http://influxdb0.${::domain}:8086", "http://influxdb1.${::domain}:8086" ],
            'database' => 'telegraf',
            'username' => 'telegraf',
            'password' => 'metricsmetricsmetrics',
    inputs   => {
        'cpu' => {
            'percpu'   => true,
            'totalcpu' => true,

Or here's a Hiera-based example (which is the recommended approach):

  role: "%{::role}"
  hostgroup: "%{::hostgroup}"
  domain: "%{::domain}"
    percpu: true
    totalcpu: true
      - "http://influxdb0.%{::domain}:8086"
      - "http://influxdb1.%{::domain}:8086"
    database: 'influxdb'
    username: 'telegraf'
    password: 'telegraf'

telegraf::inputs accepts a hash of any inputs that you'd like to configure. However, you can also optionally define individual inputs using the telegraf::input type - this suits installations where, for example, a core module sets the defaults and other modules import it.

Example 1:

telegraf::input { 'my_exec':
  plugin_type => 'exec',
  options     => {
    'commands'    => ['/usr/local/bin/',],
    'name_suffix' => '_my_input',
    'data_format' => 'json',
  require     => File['/usr/local/bin/'],

Will create the file /etc/telegraf/telegraf.d/my_exec.conf:

  commands = ['/usr/local/bin/']
  name_suffix = '_my_input'
  data_format = 'json'

Example 2:

telegraf::input { 'influxdb-dc':
  plugin_type => 'influxdb',
  options     => {
    'urls' => ['http://remote-dc:8086',],

Will create the file /etc/telegraf/telegraf.d/influxdb-dc.conf:

  urls = ["http://remote-dc:8086"]

Example 3:

telegraf::input { 'my_snmp':
  plugin_type    => 'snmp',
  options        => {
    'interval' => '60s',
  sections       => {
    '' => {
      'address'   => 'snmp_host1:161',
      'community' => 'read_only',
      'version'   => 2,
      'get_oids'  => ['',],
  single_section => {
    'snmp.tags' => {
      'environment' => 'development',

Will create the file /etc/telegraf/telegraf.d/snmp.conf:

  interval = "60s"

  address = "snmp_host1:161"
  community = "read_only"
  version = 2
  get_oids = [""]

  environment = "development"

Example 4:

class { '::telegraf':
    ensure              => '1.0.1',
    hostname            => $::hostname,
    windows_package_url => http://internal_repo:8080/chocolatey,

Will install telegraf version 1.0.1 on Windows using an internal chocolatey repo

Hierarchical configuration from multiple files

Hiera YAML and JSON backends support deep hash merging which is needed for inheriting configuration from multiple files.

First of all, make sure that the deep_merge gem is installed on your Puppet Master.

An example of hiera.yaml:

    - "roles/%{role}"
    - "type/%{virtual}"
    - "domain/%{domain}"
    - "os/%{osfamily}"
    - "common"
    - yaml
    :datadir: /etc/puppet/hiera
:merge_behavior: deeper

Then you can define configuration shared for all physical servers and place it into type/physical.yaml:

    percpu: true
    totalcpu: true

Specific roles will include some extra plugins, e.g. role/frontend.yaml:

    urls: ["http://localhost/server_status"]


This module has been developed and tested against:

  • Ubuntu 14.04
  • Debian 8
  • CentOS / RHEL 6
  • CentOS / RHEL 7
  • Windows 2008, 2008 R2, 2012, 2012 R2

Support for other distributions / operating systems is planned. Feel free to assist with development in this regard!

The configuration generated with this module is only compatible with newer releases of Telegraf, i.e 0.11.x. It won't work with the 0.2.x series.


Please fork this repository, hack away on your branch, run the tests:

$ bundle exec rake test acceptance

And then submit a pull request. Succinct, well-described and atomic commits preferred.