One place for hosting & domains

      Puppet

      Gerenciamento de Configuração 101: Escrevendo Manifests do Puppet


      Introdução

      Em poucas palavras, o gerenciamento de configuração de servidor (também conhecido como IT Automation) é uma solução para transformar a sua administração de infraestrutura em uma base de código, descrevendo todos os processos necessários para fazer o deploy de um servidor em um conjunto de scripts de provisionamento que podem ser versionados e facilmente reutilizados. Isso pode melhorar muito a integridade de qualquer infraestrutura de servidor ao longo do tempo.

      Em um guia anterior, falamos sobre os principais benefícios da implementação de uma estratégia de gerenciamento de configuração para a infraestrutura de servidor, como as ferramentas de gerenciamento de configuração funcionam e o que essas ferramentas têm em comum normalmente.

      Esta parte da série o guiará pelo processo de automatização do provisionamento de servidores usando o Puppet, uma ferramenta popular de gerenciamento de configuração capaz de gerenciar infraestruturas complexas de maneira transparente, usando um servidor mestre ou master para orquestrar a configuração dos nodes. Vamos nos concentrar na terminologia da linguagem, sintaxe e recursos necessários para criar um exemplo simplificado para automatizar completamente o deployment de um servidor web Ubuntu 18.04 usando o Apache.

      Esta é a lista de passos que precisamos automatizar para alcançar nosso objetivo:

      1. Atualizar o cache do apt
      2. Instalar o Apache
      3. Criar um diretório raiz de documentos personalizado (document root)
      4. Colocar um arquivo index.html no document root personalizado
      5. Aplicar um modelo para configurar nosso virtual host personalizado
      6. Reiniciar o Apache

      Começaremos examinando a terminologia usada pelo Puppet, seguida de uma visão geral dos principais recursos da linguagem que podem ser usados para escrever manifests. No final deste guia, compartilharemos o exemplo completo para que você possa experimentar sozinho.

      Nota: este guia pretende apresentar a linguagem do Puppet e como escrever manifests para automatizar o provisionamento de servidor. Para uma visão mais introdutória do Puppet, incluindo as etapas necessárias para instalar e começar com esta ferramenta, consulte a documentação oficial do Puppet.

      Começando

      Antes de podermos avançar para uma visão mais prática do Puppet, é importante nos familiarizarmos com importantes terminologias e conceitos introduzidos por esta ferramenta.

      Termos do Puppet

      • Puppet Master: o servidor principal que controla a configuração nos nodes
      • Puppet Agent Node: um node controlado por um Puppet Master
      • Manifest: um arquivo que contém um conjunto de instruções a serem executadas
      • Resource (recurso): uma parte do código que declara um elemento do sistema e como seu estado deve ser alterado. Por exemplo, para instalar um pacote, precisamos definir um recurso package e garantir que seu estado esteja definido como installed
      • Module: uma coleção de manifests e outros arquivos relacionados organizados de uma forma predefinida para facilitar o compartilhamento e a reutilização de partes de um provisionamento
      • Class: Assim como nas linguagens de programação regulares, as classes são usadas no Puppet para organizar melhor o provisionamento e facilitar a reutilização de partes do código
      • Facts: variáveis globais que contêm informações sobre o sistema, como interfaces de rede e sistema operacional
      • Services: usado para disparar alterações de status do serviço, como reiniciar ou parar um serviço

      O provisionamento no Puppet é escrito usando uma DSL (linguagem específica de domínio) personalizada baseada em Ruby.

      Recursos

      Com o Puppet, as tarefas ou etapas são definidas declarando recursos. Os recursos podem representar pacotes, arquivos, serviços, usuários e comandos. Eles podem ter um estado, que acionará uma alteração no sistema caso o estado de um recurso declarado seja diferente do que está atualmente no sistema. Por exemplo, um recurso package definido como installed no seu manifest acionará uma instalação do pacote no sistema se o pacote não tiver sido instalado anteriormente.

      É assim que um recurso package se parece:

      package { 'nginx':
          ensure  => 'installed'
      }
      

      Você pode executar qualquer comando arbitrário declarando um recurso exec, como o seguinte:

      exec { 'apt-get update':
          command => '/usr/bin/apt-get update'
      }
      

      Observe que a parte apt-get update na primeira linha não é a declaração de comando real, mas um identificador para este recurso exclusivo. Frequentemente, precisamos fazer referência a outros recursos de dentro de um recurso e usamos o identificador deles para isso. Nesse caso, o identificador é apt-get update, mas poderia ser qualquer outra string.

      Dependência de Recurso

      Ao escrever manifests, é importante ter em mente que o Puppet não avalia os recursos na mesma ordem em que eles são definidos. Essa é uma fonte comum de confusão para quem está começando com o Puppet. Os recursos devem definir explicitamente a dependência entre si, caso contrário, não há garantia de qual recursos será avaliado e, consequentemente, executado primeiro.

      Como um exemplo simples, digamos que você queira executar um comando, mas você precisa garantir que uma dependência seja instalada primeiro:

      package { 'python-software-properties':
          ensure => 'installed'
      }
      
      exec { 'add-repository':
          command => '/usr/bin/add-apt-repository ppa:ondrej/php5 -y'
          require => Package['python-software-properties']
      }
      

      A opção require recebe como parâmetro uma referência a outro recurso. Neste caso, estamos nos referindo ao recurso Package identificado como python-software-properties.

      É importante notar que, enquanto usamos exec, package, e assim por diante para declarar recurso (com letras minúsculas), quando nos referimos a recurso definidos anteriormente, usamos Exec, Package e assim por diante (em maiúsculas).

      Agora, digamos que você precise garantir que uma tarefa seja executada antes de outra. Para um caso como este, podemos usar a opção before:

      package { 'curl':
          ensure => 'installed'
          before => Exec['install script']
      }
      
      exec { 'install script':
          command => '/usr/bin/curl http://example.com/some-script.sh'
      

      Formato do Manifest

      Os manifests são basicamente uma coleção de declarações de recursos, usando a extensão .pp. Abaixo você pode encontrar um exemplo de um playbook simples que executa duas tarefas: atualiza o cache do apt e instala o vim posteriormente:

      exec { 'apt-get update':
          command => '/usr/bin/apt-get update'
      }
      
      package { 'vim':
          ensure => 'installed'
          require => Exec['apt-get update']
      }
      

      Antes do final deste guia, veremos um exemplo mais real de um manifest, explicado em detalhes. A próxima seção fornecerá uma visão geral dos elementos e recursos mais importantes que podem ser usados para escrever manisfests do Puppet.

      Escrevendo Manifests

      Trabalhando com Variáveis

      As variáveis podem ser definidas em qualquer ponto em um manifest. Os tipos mais comuns de variáveis são strings e matrizes ou arrays de strings, mas outros tipos também são suportados, como booleanos e hashes.

      O exemplo abaixo define uma variável string que é usada posteriormente dentro de um recurso:

      $package = "vim"
      
      package { $package:
         ensure => "installed"
      }
      

      Usando Loops

      Os loops são normalmente usados para repetir uma tarefa usando diferentes valores de entrada. Por exemplo, em vez de criar 10 tarefas para instalar 10 pacotes diferentes, você pode criar uma única tarefa e usar um loop para repetir a tarefa com todos os pacotes diferentes que deseja instalar.

      A maneira mais simples de repetir uma tarefa com valores diferentes no Puppet é usando arrays, como no exemplo abaixo:

      $packages = ['vim', 'git', 'curl']
      
      package { $packages:
         ensure => "installed"
      }
      

      A partir da versão 4, o Puppet suporta maneiras adicionais de iterar as tarefas. O exemplo abaixo faz a mesma coisa que o exemplo anterior, mas desta vez usando o iterador each. Esta opção oferece mais flexibilidade para fazer um loop pelas definições de recursos:

      $packages.each |String $package| {
        package { $package:
          ensure => "installed"
        }
      }
      

      Usando Condicionais

      Condicionais podem ser usadas para decidir dinamicamente se um bloco de código deve ou não ser executado, com base em uma variável ou em uma saída de um comando, por exemplo.

      O Puppet suporta a maioria das estruturas condicionais que você pode encontrar nas linguagens de programação tradicionais, como as instruções if/else e case. Além disso, alguns recursos como o exec suportam atributos que funcionam como condicionais, mas aceitam apenas uma saída de comando como condição.

      Digamos que você queira executar um comando com base em um fact. Nesse caso, como você deseja testar o valor de uma variável, é necessário usar uma das estruturas condicionais suportadas, como if/else:

      if $osfamily != 'Debian' {
       warning('This manifest is not supported on this OS.')
      }
      else {
       notify { 'Good to go!': }
      }
      

      Outra situação comum é quando você deseja condicionar a execução de um comando com base na saída de outro comando. Para casos como esse, você pode usar onlyif ou unless, como no exemplo abaixo. Este comando será executado apenas quando a saída de /bin/which php for bem-sucedida, ou seja, o comando sair com o status 0:

      exec { "Test":
       command => "/bin/echo PHP is installed here > /tmp/test.txt",
       onlyif => "/bin/which php"
      }
      

      Da mesma forma, unless irá executar o comando o tempo todo, exceto quando o comando sob unless sair com êxito:

      exec { "Test":
       command => "/bin/echo PHP is NOT installed here > /tmp/test.txt",
       unless => "/bin/which php"
      }
      

      Trabalhando com Templates

      Os templates geralmente são usados para definir arquivos de configuração, permitindo o uso de variáveis e outros recursos destinados a tornar esses arquivos mais versáteis e reutilizáveis. O Puppet suporta dois formatos diferentes para templates: Embedded Puppet (EPP) e Embedded Ruby (ERB). O formato EPP, no entanto, funciona apenas com versões recentes do Puppet (a partir da versão 4.0).

      Abaixo está um exemplo de um template ERB para configurar um virtual host no Apache, usando uma variável para configurar o document root para este host:

      <VirtualHost *:80>
          ServerAdmin webmaster@localhost
          DocumentRoot <%= @doc_root %>
      
          <Directory <%= @doc_root %>>
              AllowOverride All
              Require all granted
          </Directory>
      </VirtualHost>
      

      Para aplicar o template, precisamos criar um recurso file que processe o conteúdo do template com o método template. É assim que você aplicaria esse template para substituir o virtual host padrão do Apache:

      file { "/etc/apache2/sites-available/000-default.conf":
          ensure => "present",
          content => template("apache/vhost.erb") 
      }    
      

      O Puppet faz algumas suposições ao lidar com arquivos locais, a fim de reforçar a organização e a modularidade. Nesse caso, o Puppet procuraria um arquivo de template vhost.erb dentro de uma pasta apache/templates, dentro do seu diretório de módulos.

      Definindo e Acionando Serviços

      Os recursos de serviço são usados para garantir que os serviços sejam inicializados e ativados. Eles também são usados para acionar a reinicialização do serviço.

      Vamos levar em consideração o nosso exemplo anterior de uso de template, onde configuramos um virtual host no Apache. Se você deseja garantir que o Apache seja reiniciado após uma mudança no virtual host, primeiro crie um recurso service para o serviço Apache. É assim que esse recurso é definido no Puppet:

      service { 'apache2':
          ensure => running,
          enable => true
      }
      

      Agora, ao definir o recurso, você precisa incluir uma opção notify para acionar uma reinicialização:

      file { "/etc/apache2/sites-available/000-default.conf":
          ensure => "present",
          content => template("vhost.erb"),
          notify => Service['apache2'] 
      } 
      

      Manifest de Exemplo

      Agora, vamos dar uma olhada em um manifest que automatizará a instalação de um servidor Web Apache em um sistema Ubuntu 14.04, conforme discutido na introdução deste guia.

      O exemplo completo, incluindo o arquivo de template para configurar o Apache e um arquivo HTML para ser servido pelo servidor web, pode ser encontrado no Github. A pasta também contém um arquivo Vagrant que permite testar o manifest em uma configuração simplificada, usando uma máquina virtual gerenciada pelo Vagrant.

      Abaixo você pode encontrar o manifest completo:

      default.pp

      • $doc_root = "/var/www/example"
      • exec { 'apt-get update':
      • command => '/usr/bin/apt-get update'
      • }
      • package { 'apache2':
      • ensure => "installed",
      • require => Exec['apt-get update']
      • }
      • file { $doc_root:
      • ensure => "directory",
      • owner => "www-data",
      • group => "www-data",
      • mode => 644
      • }
      • file { "$doc_root/index.html":
      • ensure => "present",
      • source => "puppet:///modules/main/index.html",
      • require => File[$doc_root]
      • }
      • file { "/etc/apache2/sites-available/000-default.conf":
      • ensure => "present",
      • content => template("main/vhost.erb"),
      • notify => Service['apache2'],
      • require => Package['apache2']
      • }
      • service { 'apache2':
      • ensure => running,
      • enable => true
      • }

      Manifest Explicado

      linha 1

      O manifest começa com uma definição de variável, $doc_root. Essa variável é usada posteriormente em uma declaração de recurso.

      linhas 3-5

      Este recurso exec executa um comando apt-get update.

      linhas 7-10

      Este recurso package instala o pacote apache2, definindo que o recurso apt-get update é um requisito, o que significa que ele só será executado após a avaliação do recurso necessário.

      linhas 12-17

      Utilizamos um recurso file aqui para criar um novo diretório que servirá como document root. O recurso file pode ser usado para criar diretórios e arquivos, e também é usado para aplicar templates e copiar arquivos locais no servidor remoto. Esta tarefa pode ser executada em qualquer ponto do provisionamento, portanto, não precisamos definir nenhum require aqui.

      linhas 19-23

      Usamos outro recurso file aqui, desta vez para copiar nosso arquivo index.html local para o document root dentro do servidor. Usamos o parâmetro source para permitir que o Puppet saiba onde encontrar o arquivo original. Essa nomenclatura é baseada na maneira como o Puppet lida com arquivos locais; se você der uma olhada no repositório de exemplo do Github, você verá como a estrutura de diretórios deve ser criada para permitir que o Puppet encontre esse recurso. O diretório document root precisa ser criado antes da execução do recurso, e é por isso que incluímos uma opção require que faz referência ao recurso anterior.

      linhas 25-30

      Um novo recurso file é usado para aplicar o template do Apache e notificar o serviço para uma reinicialização. Neste exemplo, nosso provisionamento é organizado em um módulo chamado main, e é por isso que a origem do template é main/vhost.erb. Usamos uma instrução require para garantir que o recurso do template seja executado apenas após a instalação do pacote apache2, caso contrário, a estrutura de diretórios usada pelo Apache ainda não estará presente.

      linhas 32-35

      Finalmente, o recurso service declara o serviço apache2, que notificamos para reiniciar a partir do recurso que aplica o template de virtual host.

      Conclusão

      O Puppet é uma poderosa ferramenta de gerenciamento de configuração que usa uma DSL expressiva e personalizada para gerenciar recursos do servidor e automatizar tarefas. Sua linguagem oferece recursos avançados que podem dar flexibilidade extra às suas configurações de provisionamento; é importante lembrar que os recursos não são avaliados na mesma ordem em que são definidos e, por esse motivo, você precisa ter cuidado ao definir dependências entre os recursos para estabelecer a sequência de execução correta.

      No próximo guia desta série, veremos o Chef, outra ferramenta poderosa de gerenciamento de configuração que aproveita a linguagem de programação Ruby para automatizar a administração e o provisionamento da infraestrutura.



      Source link

      Install and Manage MySQL Databases with Puppet Hiera on Ubuntu 18.04


      Updated by Linode Contributed by Linode

      Puppet is a configuration management system that helps simplify the use and deployment of different types of software, making system administration more reliable and replicable. In this guide, we use Puppet to manage an installation of MySQL, a popular relational database used for applications such as WordPress, Ruby on Rails, and others. Hiera is a method of defining configuration values that Puppet will use to simplify MySQL configuration.

      In this guide, you’ll use Puppet to deploy modules on your server. At the end, you will have MySQL installed, configured, and ready to use for a variety of applications that require a database backend.

      Note

      This guide is written for a non-root user. Commands that require elevated privileges are prefixed with sudo. If you’re not familiar with the sudo command, see the Users and Groups guide.

      Before You Begin

      1. A Linode 1GB plan should be sufficient to run MySQL. Consider using a larger plan if you plan to use MySQL heavily, or for more than just a simple personal website.

      2. Familiarize yourself with our Getting Started guide and complete the steps for setting your Linode’s hostname and timezone.

      3. This guide will use sudo wherever possible. Complete the sections of our Securing Your Server to create a standard user account, harden SSH access and remove unnecessary network services.

      4. Update your system:

        sudo apt-get update && sudo apt-get upgrade
        

      Install and Configure Puppet

      Follow these steps to set up Puppet for single-host, local-only deployment. If you need to configure more than one server or to deploy a Puppet master, follow our multi-server Puppet guide.

      Install the Puppet Package

      1. Install the puppetlabs-release-bionic repository to add the Puppet packages:

        wget https://apt.puppetlabs.com/puppet-release-bionic.deb
        sudo dpkg -i puppet-release-bionic.deb
        
      2. Update the apt package index to make the Puppet Labs repository packages available, then install Puppet. This will install the puppet-agent package, which provides the puppet executable within in a compatible Ruby environment:

        sudo apt update && sudo apt install puppet-agent
        
      3. Confirm the version of Puppet installed:

        puppet --version
        

        At the time of writing, the Puppet version is 6.1.0.

      Install the Puppet MySQL Module

      Puppet Forge is a collection of modules that aid in the installation of different types of software. The MySQL module handles the installation and configuration of MySQL without you needing to manage various configuration files and services by hand.

      1. Install the MySQL module:

        sudo puppet module install puppetlabs-mysql --version 7.0.0
        

        This will install the mysql module into the default path: /etc/puppetlabs/code/environments/production/modules/.

      Puppet MySQL Manifest

      This guide uses a Puppet manifest to provide Puppet with installation and configuration instructions. Alternatively, you can configure a Puppet master.

      While the entirety of a Puppet manifest can contain the desired configuration for a host, values for Puppet classes or types can also be defined in a Hiera configuration file to simplify writing Puppet manifests in most cases. In this example, the mysql::server class parameters will be defined in Hiera, but the class must first be applied to the host.

      To apply the mysql::server class to all hosts by default, create the following Puppet manifest:

      /etc/puppetlabs/code/environments/production/manifests/site.pp
      1
      
      include ::mysql::server

      Note that site.pp is the default manifest file. Without a qualifying node { .. } line, this applies the class to any host applying the manifest. Puppet now knows to apply the mysql::server class, but still needs values for resources like databases, users, and other settings. Configure Hiera to provide these values in the next section.

      Install and Configure Puppet Hiera

      To understand how Hiera works, consider this excerpt from the default hiera.yaml file:

      /etc/puppetlabs/code/environments/production/hiera.yaml
      1
      2
      3
      4
      5
      6
      7
      
      ---
      version: 5
      hierarchy:
        - name: "Per-node data"
          path: "nodes/%{::trusted.certname}.yaml"
        - name: "Common data"
          path: "common.yaml"

      This Hiera configuration instructs Puppet to accept variable values from nodes/%{::trusted.certname}.yaml. If your Linode’s hostname is examplehostname, define a file called nodes/examplehostname.yaml). Any variables found in YAML files higher in the hierarchy are preferred, while any variable names that do not exist in those files will fall-through to files lower in the hierarchy (in this example, common.yaml).

      The following configuration will define Puppet variables in common.yaml to inject variables into the mysql::server class.

      Initial Hiera Configuration

      Hiera configuration files are formatted as yaml, with keys defining the Puppet parameters to inject their associated values. To get started, set the MySQL root password. The following example of a Puppet manifest is one way to control this password:

      example.pp
      1
      2
      3
      
      class { '::mysql::server':
        root_password => 'examplepassword',
      }

      We can also define the root password with the following Hiera configuration file. Create the following YAML file and note how the root_password parameter is defined as Hiera yaml:

      /etc/puppetlabs/code/environments/production/data/common.yaml
      1
      
      mysql::server::root_password: examplepassword

      Replace examplepassword with the secure password of your choice. Run Puppet to set up MySQL with default settings and the chosen root password:

      sudo -i puppet apply /etc/puppetlabs/code/environments/production/manifests/site.pp
      

      Puppet will output its progress before completing. To confirm MySQL has been configured properly, run a command:

      mysql -u root -p -e 'select version();'
      

      Enter the password and MySQL returns its version:

      +-------------------------+
      | version()               |
      +-------------------------+
      | 5.7.24-0ubuntu0.18.04.1 |
      +-------------------------+
      

      Define MySQL Resources

      Using Hiera, we can define the rest of the MySQL configuration entirely in yaml. The following steps will create a database and user for use in a WordPress installation.

      1. Create a pre-hashed MySQL password. Replace the password wordpresspassword in this example, and when prompted for a the root MySQL password, use the first root password chosen in the previous section to authenticate. Note the string starting with a * that the command returns for Step 2:

        mysql -u root -p -NBe 'select password("wordpresspassword")'
        *E62D3F829F44A91CC231C76347712772B3B9DABC
        
      2. With the MySQL password hash ready, we can define Hiera values. The following YAML defines parameters to create a database called wordpress and a user named wpuser that has permission to connect from localhost. The YAML also defines a GRANT allowing wpuser to operate on the wordpress database with ALL permissions:

        /etc/puppetlabs/code/environments/production/data/common.yaml
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        
        mysql::server::root_password: examplepassword
        mysql::server::databases:
          wordpress:
            ensure: present
        mysql::server::users:
          wpuser@localhost:
            ensure: present
            password_hash: '*E62D3F829F44A91CC231C76347712772B3B9DABC'
        mysql::server::grants:
          wpuser@localhost/wordpress.*:
            ensure: present
            privileges: ALL
            table: wordpress.*
            user: wpuser@localhost
      3. Re-run Puppet:

        sudo -i puppet apply /etc/puppetlabs/code/environments/production/manifests/site.pp
        
      4. The wpuser should now be able to connect to the wordpress database. To verify, connect to the MySQL daemon as the user wpuser to the wordpress database:

        mysql -u wpuser -p wordpress
        

        After you enter the password for wpuser, exit the MySQL prompt:

        exit
        

      Add Hierarchies for Specific Environments

      Additional configurations can be added that will only be applied to specific environments. For example, backup jobs may only be applied for hosts in a certain region, or specific databases can be created in a particular deployment.

      In the following example, Puppet will configure the MySQL server with one additional database, but only if that server’s distribution is Debian-based.

      1. Modify hiera.yaml to contain the following:

        /etc/puppetlabs/code/environments/production/hiera.yaml
        1
        2
        3
        4
        5
        6
        7
        8
        
        ---
        version: 5
        hierarchy:
          - name: "Per OS Family"
            path: "os/%{facts.os.family}.yaml"
          - name: "Other YAML hierarchy levels"
            paths:
              - "common.yaml"

        This change instructs Hiera to look for Puppet parameters first in "os/%{facts.os.family}.yaml" and then in common.yaml. The first, fact-based element of the hierarchy is dynamic, and dependent upon the host that Puppet and Hiera control. In this Ubuntu-based example, Hiera will look for Debian.yaml in the os folder, while on a distribution such as CentOS, the file RedHat.yaml will automatically be referenced instead.

      2. Create the following YAML file:

        /etc/puppetlabs/code/environments/production/data/os/Debian.yaml
        1
        2
        3
        4
        5
        6
        7
        
        lookup_options:
          mysql::server::databases:
            merge: deep
        
        mysql::server::databases:
          ubuntu-backup:
            ensure: present

        Though similar to the common.yaml file defined in previous steps, this file will add the ubuntu-backup database only on Debian-based hosts (like Ubuntu). In addition, the lookup_options setting ensures that the mysql::server:databases parameter is merged between Debian.yaml and common.yaml so that all databases are managed. Without lookup_options set to deeply merge these hashes, only the most specific hierarchy file will be applied to the host, in this case, Debian.yaml.

        • Alternatively, because our Puppet manifest is short, we can test the same command using the -e flag to apply an inline manifest:

          sudo -i puppet apply -e 'include ::mysql::server'
          
      3. Run Puppet and observe the changes:

        sudo -i puppet apply /etc/puppetlabs/code/environments/production/manifests/site.pp
        
      4. Verify that the new database exists:

        mysql -u root -p -e 'show databases;'
        

        This includes the new ubuntu-backup database:

        +---------------------+
        | Database            |
        +---------------------+
        | information_schema  |
        | mysql               |
        | performance_schema  |
        | sys                 |
        | ubuntu-backup       |
        | wordpress           |
        +---------------------+
        

      Congratulations! You can now control your Puppet configuration via highly configurable Hiera definitions.

      More Information

      You may wish to consult the following resources for additional information on this topic. While these are provided in the hope that they will be useful, please note that we cannot vouch for the accuracy or timeliness of externally hosted materials.

      Find answers, ask questions, and help others.

      This guide is published under a CC BY-ND 4.0 license.



      Source link

      Getting Started with Puppet – Basic Installation and Setup


      Updated by Linode Written by Linode

      Puppet is a configuration management tool that simplifies system administration. Puppet uses a client/server model in which your managed nodes, running a process called the Puppet agent, talk to and pull down configuration profiles from a Puppet master.

      Puppet deployments can range from small groups of servers up to enterprise-level operations. This guide will demonstrate how to install Puppet 6.1 on three servers:

      • A Puppet master running Ubuntu 18.04
      • A managed Puppet node running Ubuntu 18.04
      • A managed Puppet node running CentOS 7

      After installation, the next section will show you how to secure these servers via Puppet. This section will demonstrate core features of the Puppet language.

      Note

      Most guides will instruct you to follow the How to Secure your Server guide before proceeding. Because Puppet will be used to perform this task, you should begin this guide as the root user. A limited user with administrative privileges will be configured via Puppet in later steps.

      Before You Begin

      The following table displays example system information for the servers that will be deployed in this guide:

      Description OS Hostname FQDN IP
      Puppet master Ubuntu 18.04 puppet puppet.example.com 192.0.2.2
      Node 1 (Ubuntu) Ubuntu 18.04 puppet-agent-ubuntu puppet-agent-ubuntu.example.com 192.0.2.3
      Node 2 (CentOS) CentOS 7 puppet-agent-centos puppet-agent-centos.example.com 192.0.2.4

      You can choose different hostnames and fully qualified domain names (FQDN) for each of your servers, and the IP addresses for your servers will be different from the example addresses listed. You will need to have a registered domain name in order to specify FQDNs for your servers.

      Throughout this guide, commands and code snippets will reference the values displayed in this table. Wherever such a value appears, replace it with your own value.

      Create your Linodes

      1. Create three Linodes corresponding to the servers listed in the table above. Your Puppet master Linode should have at least four CPU cores; the Linode 8GB plan is recommended. The two other nodes can be of any plan size, depending on how you intend to use them after Puppet is installed and configured.

      2. Configure your timezone on your master and agent nodes so that they all have the same time data.

      3. Set the hostname for each server.

      4. Set the FQDN for each Linode by editing the servers’ /etc/hosts files.

        Example content for the hosts file

        You can model the contents of your /etc/hosts files on these snippets:

        Master
        1
        2
        3
        4
        5
        6
        7
        
        127.0.0.1   localhost
        192.0.2.2   puppet.example.com puppet
        
        # The following lines are desirable for IPv6 capable hosts
        ::1     localhost ip6-localhost ip6-loopback
        ff02::1 ip6-allnodes
        ff02::2 ip6-allrouters
        Node 1 (Ubuntu)
        1
        2
        3
        4
        5
        6
        7
        
        127.0.0.1   localhost
        192.0.2.3   puppet-agent-ubuntu.example.com puppet-agent-ubuntu
        
        # The following lines are desirable for IPv6 capable hosts
        ::1     localhost ip6-localhost ip6-loopback
        ff02::1 ip6-allnodes
        ff02::2 ip6-allrouters
        Node 2 (CentOS)
        1
        2
        3
        
        127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
        ::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
        192.0.2.4   puppet-agent-centos.example.com puppet-agent-centos
      5. Set up DNS records for your Linodes’ FQDNs. For each Linode, create a new A record with the name specified by its FQDN and assign it to that Linode’s IP address.

        If you don’t use Linode’s name servers for your domain, consult your name server authority’s website for instructions on how to edit your DNS records.

        Updating DNS records at common nameserver authorities

        The following support documents describe how to update DNS records at common nameserver authorities:

      Puppet Master

      Install the Puppet Server Software

      The Puppet master runs the puppetserver service, which is responsible for compiling and supplying configuration profiles to your managed nodes.

      The puppetserver service has the Puppet agent service as a dependency (which is just called puppet when running on your system). This means that the agent software will also be installed and can be run on your master. Because your master can run the agent service, you can configure your master via Puppet just as you can configure your other managed nodes.

      1. Log in to your Puppet master via SSH (as root):

        ssh root@puppet.example.com
        
      2. Download the Puppet repository, update your system packages, and install puppetserver:

        wget https://apt.puppetlabs.com/puppet-release-bionic.deb
        dpkg -i puppet-release-bionic.deb
        apt update
        apt install puppetserver
        

      Configure the Server Software

      1. Use the puppet config command to set values for the dns_alt_names setting:

        /opt/puppetlabs/bin/puppet config set dns_alt_names 'puppet,puppet.example.com' --section main
        

        If you inspect the configuration file, you’ll see that the setting has been added:

        cat /etc/puppetlabs/puppet/puppet.conf
        
          
        [main]
        dns_alt_names = puppet,puppet.example.com
        # ...
        
        

        Note

        The puppet command by default is not added to your PATH. Using Puppet’s interactive commands requires a full file path. To avoid this, update your PATH for your existing shell session:

        export PATH=/opt/puppetlabs/bin:$PATH
        

        A more permanent solution would be to add this to your .profile or .bashrc files.

      2. Update your Puppet master’s /etc/hosts to resolve your managed nodes’ IP addresses. For example, your /etc/hosts file might look like the following:

        /etc/hosts
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        
        127.0.0.1   localhost
        192.0.2.2   puppet.example.com puppet
        
        192.0.2.3   puppet-agent-ubuntu.example.com puppet-agent-ubuntu
        192.0.2.4   puppet-agent-centos.example.com puppet-agent-centos
        
        # The following lines are desirable for IPv6 capable hosts
        ::1     localhost ip6-localhost ip6-loopback
        ff02::1 ip6-allnodes
        ff02::2 ip6-allrouters

        Note

      3. Start and enable the puppetserver service:

        systemctl start puppetserver
        systemctl enable puppetserver
        

        By default, the Puppet master listens for client connections on port 8140. If the puppetserver service fails to start, check that the port is not already in use:

        netstat -anpl | grep 8140
        

      Puppet Agents

      Install Puppet Agent

      1. On your managed node running Ubuntu 18.04, install the puppet-agent package:

        wget https://apt.puppetlabs.com/puppet-release-bionic.deb
        dpkg -i puppet-release-bionic.deb
        apt update
        apt install puppet-agent
        
      2. On your managed node running CentOS 7, enter:

        rpm -Uvh https://yum.puppet.com/puppet/puppet-release-el-7.noarch.rpm
        yum install puppet-agent
        

      Configure Puppet Agent

      1. Modify your managed nodes’ hosts files to resolve the Puppet master’s IP. To do so, add a line like:

        /etc/hosts
        1
        
        192.0.2.2    puppet.example.com puppet

        Example content for the hosts file

        You can model the contents of your managed nodes’ /etc/hosts files on the following snippets. These incorporate the FQDN declarations described in the Create your Linodes section:

        Node 1 (Ubuntu)
        1
        2
        3
        4
        5
        6
        7
        8
        9
        
        127.0.0.1   localhost
        192.0.2.3   puppet-agent-ubuntu.example.com puppet-agent-ubuntu
        
        192.0.2.2   puppet.example.com puppet
        
        # The following lines are desirable for IPv6 capable hosts
        ::1     localhost ip6-localhost ip6-loopback
        ff02::1 ip6-allnodes
        ff02::2 ip6-allrouters
        Node 2 (CentOS)
        1
        2
        3
        4
        5
        
        127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
        ::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
        192.0.2.4   puppet-agent-centos.example.com puppet-agent-centos
        
        192.0.2.2   puppet.example.com puppet
      2. On each managed node, use the puppet config command to set the value for your server setting to the FQDN of the master:

        /opt/puppetlabs/bin/puppet config set server 'puppet.example.com' --section main
        

        If you inspect the configuration file on the nodes, you’ll see that the setting has been added:

        cat /etc/puppetlabs/puppet/puppet.conf
        
          
        [main]
        server = puppet.example.com
        # ...
        
        
      3. Use the puppet resource command to start and enable the Puppet agent service:

        /opt/puppetlabs/bin/puppet resource service puppet ensure=running enable=true
        

        Note

        On systemd systems, the above command is equivalent to using these two systemctl commands:

        systemctl start puppet
        systemctl enable puppet
        

      Generate and Sign Certificates

      Before your managed nodes can receive configurations from the master, they first need to be authenticated:

      1. On your Puppet agents, generate a certificate for the Puppet master to sign:

        /opt/puppetlabs/bin/puppet agent -t
        

        This command will output an error, stating that no certificate has been found. This error is because the generated certificate needs to be approved by the Puppet master.

      2. Log in to your Puppet master and list the certificates that need approval:

        /opt/puppetlabs/bin/puppetserver ca list
        

        It should output a list with your agent nodes’ hostnames.

      3. Approve the certificates:

        /opt/puppetlabs/bin/puppetserver ca sign --certname puppet-agent-ubuntu.example.com,puppet-agent-centos.example.com
        
      4. Return to the Puppet agent nodes and run the Puppet agent again:

        /opt/puppetlabs/bin/puppet agent -t
        

        You should see something like the following:

          
        Info: Downloaded certificate for hostname.example.com from puppet
        Info: Using configured environment 'production'
        Info: Retrieving pluginfacts
        Info: Retrieving plugin
        Info: Retrieving locales
        Info: Caching catalog for hostname.example.com
        Info: Applying configuration version '1547066428'
        Info: Creating state file /opt/puppetlabs/puppet/cache/state/state.yaml
        Notice: Applied catalog in 0.02 seconds
        
        

      Add Modules to Configure Agent Nodes

      The Puppet master and agent nodes are now functional, but they are not secure. Based on concepts from the How to Secure your Server guide, a limited user and a firewall should be configured. This can be done on all nodes through the creation of basic Puppet modules, shown below.

      Note

      This is not meant to provide a basis for a fully-hardened server, and is intended only as a starting point. Alter and add firewall rules and other configuration options, depending on your specific needs.

      Puppet modules are Puppet’s prescribed way of organizing configuration code to serve specific purposes, like installing and configuration an application. You can create custom modules, or you can download and use modules published on Puppet Forge.

      Add a Limited User

      To create a new limited user on your nodes, you will create and apply a new module called accounts. This module will employ the user resource.

      1. From the Puppet master, navigate to the /etc/puppetlabs/code/environments/production/modules directory. When a managed node requests its configuration from the master, the Puppet server process will look in this location for your modules:

        cd /etc/puppetlabs/code/environments/production/modules/
        
      2. Create the directory for a new accounts module:

        mkdir accounts
        cd accounts
        
      3. Create the following directories inside the accounts module:

        mkdir {examples,files,manifests,templates}
        
        Directory Description
        manifests The Puppet code which powers the module
        files Static files to be copied to managed nodes
        templates Template files to be copied to managed nodes that can e customized with variables
        examples Example code which shows how to use the module

        Note

        Review Puppet’s Module fundamentals article for more information on how a module is structured.
      4. Navigate to the manifests directory:

        cd manifests
        
      5. Any file which contains Puppet code is called a manifest, and each manifest file ends in .pp. When located inside a module, a manifest should only define one class. If a module’s manifests directory has an init.pp file, the class definition it contains is considered the main class for the module. The class definition inside init.pp should have the same name as the module.

        Create an init.pp file with the contents of the following snippet. Replace all instances of username with a username of your choosing:

        accounts/manifests/init.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        
        class accounts {
        
          user { 'username':
            ensure      => present,
            home        => '/home/username',
            shell       => '/bin/bash',
            managehome  => true,
            gid         => 'username',
          }
        
        }
        Option Description
        ensure Ensures that the user exists if set to present, or does not exist if set to absent
        home The path for the user’s home directory
        managehome Controls whether a home directory should be created when creating the user
        shell The path to the shell for the user
        gid The user’s primary group
      6. Although the class declares what the user’s primary group should be, it will not create the group itself. Create a new file called groups.pp inside the manifests directory with the following contents. Replace username with your chosen username:

        accounts/manifests/groups.pp
        1
        2
        3
        4
        5
        6
        7
        
        class accounts::groups {
        
          group { 'username':
            ensure  => present,
          }
        
        }
      7. Your accounts class can declare your new accounts::groups class for use within the accounts class scope. Open your init.pp in your editor and enter a new include declaration at the beginning of the class:

        accounts/manifests/init.pp
        1
        2
        3
        4
        5
        6
        7
        
        class accounts {
        
          include accounts::groups
        
          # ...
        
        }
      8. The new user should have administrative privileges. Because we have agent nodes on both Debian- and Red Hat-based systems, the new user needs to be in the sudo group on Debian systems, and the wheel group on Red Hat systems.

        This value can be set dynamically through the use of Puppet facts. The facts system collects system information about your nodes and makes it available in your manifests.

        Add a selector statement to the top of your accounts class:

        accounts/manifests/init.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        
        class accounts {
        
          $rootgroup = $osfamily ? {
            'Debian'  => 'sudo',
            'RedHat'  => 'wheel',
            default   => warning('This distribution is not supported by the Accounts module'),
          }
        
          include accounts::groups
        
          # ...
        
        }

        This code defines the value for the $rootgroup variable by checking the value of $osfamily, which is one of Puppet’s core facts. If the value for $osfamily does not match Debian or Red Hat, the default value will output a warning that the distribution selected is not supported by this module.

        Note

        The Puppet Configuration Language executes code from top to bottom. Because the user resource declaration will reference the $rootgroup variable, you must define $rootgroup before the user declaration.

      9. Update the user resource to include the groups option as follows:

        accounts/manifests/init.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        
        # ...
        
        user { 'username':
          ensure      => present,
          home        => '/home/username',
          shell       => '/bin/bash',
          managehome  => true,
          gid         => 'username',
          groups      => "$rootgroup",
        }
        
        # ...

        The value "$rootgroup" is enclosed in double quotes " " instead of single quotes ' ' because it is a variable which needs to be interpolated in your code.

      10. The final value that needs to be added is the user’s password. Since we do not want to use plain text, the password should be supplied to Puppet as a SHA1 digest, which is supported by default. Generate a digest with the openssl command:

        openssl passwd -1
        

        You will be prompted to enter your password. A hashed password will be output. Copy this value to your clipboard.

      11. Update the user resource to include the password option as follows; insert your copied password hash as the value for the option:

        accounts/manifests/init.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        
        # ...
        
        user { 'username':
          ensure      => present,
          home        => '/home/username',
          shell       => '/bin/bash',
          managehome  => true,
          gid         => 'username',
          groups      => "$rootgroup",
          password    => 'your_password_hash',
        }
        
        # ...

        Caution

        The hashed password must be included in single quotes ' '.

      12. After saving your changes, use the Puppet parser to ensure that the code is correct:

        /opt/puppetlabs/bin/puppet parser validate init.pp
        

        Any errors that need to be addressed will be logged to standard output. If nothing is returned, your code is valid.

      13. Navigate to the examples directory and create another init.pp file:

        cd ../examples
        
        accounts/examples/init.pp
      14. While still in the examples directory, test the module:

        /opt/puppetlabs/bin/puppet apply --noop init.pp
        

        Note

        The --noop parameter prevents Puppet from actually applying the module to your system and making any changes.

        It should return:

          
        Notice: Compiled catalog for puppet.example.com in environment production in 0.26 seconds
        Notice: /Stage[main]/Accounts::Groups/Group[username]/ensure: current_value absent, should be present (noop)
        Notice: Class[Accounts::Groups]: Would have triggered 'refresh' from 1 events
        Notice: /Stage[main]/Accounts/User[username]/ensure: current_value absent, should be present (noop)
        Notice: Class[Accounts]: Would have triggered 'refresh' from 1 events
        Notice: Stage[main]: Would have triggered 'refresh' from 2 events
        Notice: Finished catalog run in 0.02 seconds
        
        
      15. Again from the examples directory, run puppet apply to make these changes to the Puppet master server:

        /opt/puppetlabs/bin/puppet apply init.pp
        

        Puppet will create your limited Linux user on your master.

      16. Log out as root and log in to the Puppet master as your new user.

      Edit SSH Settings

      Although a new limited user has successfully been added to the Puppet master, it is still possible to login to the system as root. To properly secure your system, root access should be disabled.

      Note

      Because you are now logged in to the Puppet master as a limited user, you will need to execute commands and edit files with the user’s sudo privileges.

      1. Navigate to the files directory within the accounts module:

        cd /etc/puppetlabs/code/environments/production/modules/accounts/files
        
      2. Copy your system’s existing sshd_config file to this directory:

        sudo cp /etc/ssh/sshd_config .
        
      3. Open the file in your editor (making sure that you open it with sudo privileges) and set the PermitRootLogin value to no:

        accounts/files/sshd_config
      4. Navigate back to the manifests directory:

        cd ../manifests
        
      5. Create a new manifest called ssh.pp. Use the file resource to replace the default SSH configuration file with one managed by Puppet:

        accounts/manifests/ssh.pp
        1
        2
        3
        4
        5
        6
        7
        8
        
        class accounts::ssh {
        
          file { '/etc/ssh/sshd_config':
            ensure  => present,
            source  => 'puppet:///modules/accounts/sshd_config',
          }
        
        }

        Note

        The files directory is omitted from the source line because the files folder is the default location of files within a module. For more information on the format used to access resources in a module, refer to the official Puppet module documentation.
      6. Create a second resource to restart the SSH service and set it to run whenever sshd_config is changed. This will also require a selector statement because the SSH service is named ssh on Debian systems and sshd on Red Hat systems:

        accounts/manifests/ssh.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        
        class accounts::ssh {
        
          $sshname = $osfamily ? {
            'Debian'  => 'ssh',
            'RedHat'  => 'sshd',
            default   => warning('This distribution is not supported by the Accounts module'),
          }
        
          file { '/etc/ssh/sshd_config':
            ensure  => present,
            source  => 'puppet:///modules/accounts/sshd_config',
            notify  => Service["$sshname"],
          }
        
          service { "$sshname":
            hasrestart  => true,
          }
        
        }

        Note

      7. Include the accounts::ssh class within the accounts class in init.pp:

        accounts/manifests/init.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        
        class accounts {
        
          # ...
        
          include accounts::groups
          include accounts::ssh
        
          # ...
        
        }

        The complete init.pp

        The contents of your init.pp should now look like the following snippet:

        accounts/manifests/init.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
        21
        22
        
        class accounts {
        
            $rootgroup = $osfamily ? {
                'Debian' => 'sudo',
                'RedHat' => 'wheel',
                default => warning('This distro not supported by Accounts module'),
            }
        
            include accounts::groups
            include accounts::ssh
        
            user { 'example':
                ensure  => present,
                home    => '/home/username',
                shell   => '/bin/bash',
                managehome  => true,
                gid     => 'username',
                groups  => "$rootgroup",
                password => 'your_password_hash'
            }
        
        }
      8. Run the Puppet parser to test the syntax of the new class, then navigate to the examples directory to test and run the update to your accounts class:

        sudo /opt/puppetlabs/bin/puppet parser validate ssh.pp
        cd ../examples
        sudo /opt/puppetlabs/bin/puppet apply --noop init.pp
        sudo /opt/puppetlabs/bin/puppet apply init.pp
        

        Note

        You may see the following line in your output when validating:

          
        Error: Removing mount "files": /etc/puppet/files does not exist or is not a directory
        
        

        This refers to a Puppet configuration file, not the module resource you’re trying to copy. If this is the only error in your output, the operation should still succeed.

      9. To ensure that the ssh class is working properly, log out of the Puppet master and then try to log in as root. You should not be able to do so.

      Add and Configure IPtables

      To complete this guide’s security settings, the firewall needs to be configure on your Puppet master and nodes. The iptables firewall software will be used.

      1. By default, changes to your iptables rules will not persist across reboots. To avoid this, install the appropriate package on your Puppet master and nodes:

        Ubuntu/Debian:

        sudo apt install iptables-persistent
        

        CentOS 7:

        CentOS 7 uses firewalld by default as a controller for iptables. Be sure firewalld is stopped and disabled before starting to work directly with iptables:

        sudo systemctl stop firewalld && sudo systemctl disable firewalld
        sudo yum install iptables-services
        
      2. On your Puppet master, install Puppet Lab’s firewall module from the Puppet Forge:

        sudo /opt/puppetlabs/bin/puppet module install puppetlabs-firewall
        

        The module will be installed in your /etc/puppetlabs/code/environments/production/modules directory.

      3. Navigate to the manifests directory inside the new firewall module:

        cd /etc/puppetlabs/code/environments/production/modules/firewall/manifests/
        
      4. Create a file titled pre.pp, which will contain all basic networking rules that should be run first:

        firewall/manifests/pre.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
        21
        22
        23
        24
        25
        26
        27
        28
        29
        30
        31
        32
        33
        34
        35
        36
        37
        38
        39
        40
        41
        42
        43
        44
        45
        46
        47
        48
        49
        50
        51
        52
        53
        54
        55
        
        class firewall::pre {
        
          Firewall {
            require => undef,
          }
        
           # Accept all loopback traffic
          firewall { '000 lo traffic':
            proto       => 'all',
            iniface     => 'lo',
            action      => 'accept',
          }->
        
           #Drop non-loopback traffic
          firewall { '001 reject non-lo':
            proto       => 'all',
            iniface     => '! lo',
            destination => '127.0.0.0/8',
            action      => 'reject',
          }->
        
           #Accept established inbound connections
          firewall { '002 accept established':
            proto       => 'all',
            state       => ['RELATED', 'ESTABLISHED'],
            action      => 'accept',
          }->
        
           #Allow all outbound traffic
          firewall { '003 allow outbound':
            chain       => 'OUTPUT',
            action      => 'accept',
          }->
        
           #Allow ICMP/ping
          firewall { '004 allow icmp':
            proto       => 'icmp',
            action      => 'accept',
          }
        
           #Allow SSH connections
          firewall { '005 Allow SSH':
            dport    => '22',
            proto   => 'tcp',
            action  => 'accept',
          }->
        
           #Allow HTTP/HTTPS connections
          firewall { '006 HTTP/HTTPS connections':
            dport    => ['80', '443'],
            proto   => 'tcp',
            action  => 'accept',
          }
        
        }
      5. In the same directory, create post.pp, which will run any firewall rules that need to be input last:

        firewall/manifests/post.pp
        1
        2
        3
        4
        5
        6
        7
        8
        9
        
        class firewall::post {
        
          firewall { '999 drop all':
            proto  => 'all',
            action => 'drop',
            before => undef,
          }
        
        }

        These rules will direct the system to drop all inbound traffic that is not already permitted in the firewall.

      6. Run the Puppet parser on both files to check their syntax for errors:

        sudo /opt/puppetlabs/bin/puppet parser validate pre.pp
        sudo /opt/puppetlabs/bin/puppet parser validate post.pp
        
      7. Navigate to the main manifests directory:

        cd /etc/puppetlabs/code/environments/production/manifests
        
      8. Create a file named site.pp inside /etc/puppetlabs/code/environments/production/manifests. This file is the main manifest for the Puppet server service. It is used to map modules, classes, and resources to the nodes that they should be applied to.

        site.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
        21
        22
        23
        24
        25
        26
        
        node default {
        
        }
        
        node 'puppet.example.com' {
        
          include accounts
        
          resources { 'firewall':
            purge => true,
          }
        
          Firewall {
            before        => Class['firewall::post'],
            require       => Class['firewall::pre'],
          }
        
          class { ['firewall::pre', 'firewall::post']: }
        
          firewall { '200 Allow Puppet Master':
            dport         => '8140',
            proto         => 'tcp',
            action        => 'accept',
          }
        
        }
      9. Run the site.pp file through the Puppet parser to check its syntax for errors. Then, test the file with the --noop option to see if it will run:

        sudo /opt/puppetlabs/bin/puppet parser validate site.pp
        sudo /opt/puppetlabs/bin/puppet apply --noop site.pp
        

        If successful, run puppet apply without the --noop option:

        sudo /opt/puppetlabs/bin/puppet apply site.pp
        
      10. Once Puppet has finished applying the changes, check the Puppet master’s iptables rules:

        sudo iptables -L
        

        It should return:

        Chain INPUT (policy ACCEPT)
        target     prot opt source               destination
        ACCEPT     all  --  anywhere             anywhere             /* 000 lo traffic */
        REJECT     all  --  anywhere             127.0.0.0/8          /* 001 reject non-lo */ reject-with icmp-port-unreachable
        ACCEPT     all  --  anywhere             anywhere             /* 002 accept established */ state RELATED,ESTABLISHED
        ACCEPT     icmp --  anywhere             anywhere             /* 004 allow icmp */
        ACCEPT     tcp  --  anywhere             anywhere             multiport ports ssh /* 005 Allow SSH */
        ACCEPT     tcp  --  anywhere             anywhere             multiport ports http,https /* 006 HTTP/HTTPS connections */
        ACCEPT     tcp  --  anywhere             anywhere             multiport ports 8140 /* 200 Allow Puppet Master */
        DROP       all  --  anywhere             anywhere             /* 999 drop all */
        
        Chain FORWARD (policy ACCEPT)
        target     prot opt source               destination
        
        Chain OUTPUT (policy ACCEPT)
        target     prot opt source               destination
        ACCEPT     tcp  --  anywhere             anywhere             /* 003 allow outbound */
        

      Apply Modules to the Agent Nodes

      Now that the accounts and firewall modules have been created, tested, and run on the Puppet master, it is time to apply them to your managed nodes.

      1. On the Puppet master, navigate to /etc/puppetlabs/code/environments/production/manifests:

        cd /etc/puppetlabs/code/environments/production/manifests
        
      2. Update site.pp to declare the modules, classes, and resources that should be applied to each managed node:

        site.pp
         1
         2
         3
         4
         5
         6
         7
         8
         9
        10
        11
        12
        13
        14
        15
        16
        17
        18
        19
        20
        21
        22
        23
        24
        25
        26
        27
        28
        29
        30
        31
        32
        33
        34
        35
        36
        37
        38
        39
        40
        41
        
        node default {
        
        }
        
        node 'puppet.example.com' {
          # ...
        }
        
        node 'puppet-agent-ubuntu.example.com' {
        
          include accounts
        
          resources { 'firewall':
            purge => true,
          }
        
          Firewall {
            before        => Class['firewall::post'],
            require       => Class['firewall::pre'],
          }
        
          class { ['firewall::pre', 'firewall::post']: }
        
        }
        
        node 'puppet-agent-centos.example.com' {
        
          include accounts
        
          resources { 'firewall':
            purge => true,
          }
        
          Firewall {
            before        => Class['firewall::post'],
            require       => Class['firewall::pre'],
          }
        
          class { ['firewall::pre', 'firewall::post']: }
        
        }
      3. By default, the Puppet agent service on your managed nodes will automatically check with the master once every 30 minutes and apply any new configurations from the master. You can also manually invoke the Puppet agent process in-between automatic agent runs.

        Log in to each managed node (as root) and run the Puppet agent:

        /opt/puppetlabs/bin/puppet agent -t
        
      4. To ensure the Puppet agent worked:

      Congratulations! You’ve successfully installed Puppet on a master and two managed nodes. Now that you’ve confirmed everything is working, you can create additional modules to automate configuration management on your nodes. For more information, review Puppet’s open source documentation. You can also install and use modules others have created on the Puppet Forge.

      More Information

      You may wish to consult the following resources for additional information on this topic. While these are provided in the hope that they will be useful, please note that we cannot vouch for the accuracy or timeliness of externally hosted materials.

      Find answers, ask questions, and help others.

      This guide is published under a CC BY-ND 4.0 license.



      Source link