Installing Puppet: From Source Tarballs

Included in Puppet Enterprise 3.8. A newer version is available; see the version menu above for details.

Note: This document covers open source releases of Puppet version 3.8 and lower. For current versions, you should see instructions for installing the latest version of Puppet or installing Puppet Enterprise.

On *nix platforms without native packages available, you can install Puppet with a tarball of source code. This is even more unwieldy than using RubyGems and is almost never recommended, but it might be necessary in some cases.

First

Before installing Puppet, make sure you’ve looked at the pre-install tasks.

Step 1: Ensure Prerequisites are Installed

  1. Use your OS’s package tools to install Ruby. In some cases, you might need to compile and install it yourself.
  2. On Linux platforms, ensure that the LSB tools are installed. We recommend installing at least lsb_release. See your OS’s documentation for details about its LSB tools.
  3. If you use Ruby 1.8.7, you must also install the JSON library. If you have RubyGems installed, you can use sudo gem install json_pure, but if not, you’ll need to download the source from the project’s site and install it manually.

Step 2: Download Puppet and its Dependencies

Download the latest .tar.gz files containing the Puppet, Facter, and Hiera sources.

Optionally, you can also download the corresponding .asc files and verify the signature of each tarball.

Step 3: Install Facter, Hiera, and Puppet

Unarchive the Facter, Hiera, and Puppet tarballs, then navigate to each resulting directory and run:

sudo ruby install.rb

Upgrading

Note: Read our tips on upgrading before upgrading your Puppet deployment.

To upgrade to the latest versions of Puppet, Facter, and Hiera, download new tarballs and run the install scripts again.

After upgrading, restart the puppet service and, if applicable, the Puppet master web server.

Step 6: Pre-Configure Puppet

Source tarballs lack some of the conveniences that native packages have. To prepare for the post-install tasks, do the following:

Create Users

The Puppet master service needs a puppet user and group. If this node might ever act as a Puppet master server, do the following:

  1. Create a puppet group:

    sudo puppet resource group puppet ensure=present
    
  2. Create a puppet user:

    sudo puppet resource user puppet ensure=present gid=puppet shell='/sbin/nologin'
    

This step is optional on agents.

Create an Init Script for Puppet Agent

To run Puppet agent as a service, you need an init script (or whatever your OS’s equivalent is).

Because each platform handles init scripts differently, you must find one specific to your operating system or build one yourself based on what you know about your chosen operating system. The ext directory in the Puppet source includes several example init scripts or service configurations for various operating systems:

Do not start the puppet or Puppet master services yet.

Create a Confdir and Config Files

When you install with tarballs, your confdir might be empty or missing. Populate it with these required files:

  1. Make sure the confdir exists. This directory is usually at /etc/puppet/.
  2. Create a puppet.conf file in the confdir. You can leave it empty until the post-install tasks.
  3. If this node will ever act as a Puppet master, it needs an auth.conf file. Use the default auth.conf file from the Puppet source, and put it in the confdir.
  4. Create a main manifest and a module directory. Alternately, enable directory environments in puppet.conf and create a production environment, as described in the directory environments page.

Make Sure Puppet Can be Loaded

Run puppet --version. This should output the version of Puppet you now have installed.

Next

At this point, Puppet is installed, but it isn’t configured or running. You should now do the post-install tasks.

↑ Back to top