Skip to content

Latest commit

 

History

History
142 lines (100 loc) · 7.31 KB

connect_puppet_apply.markdown

File metadata and controls

142 lines (100 loc) · 7.31 KB
Raw
title layout canonical
PuppetDB 3.1 » Connecting Standalone Puppet Nodes to PuppetDB
default
/puppetdb/latest/connect_puppet_apply.html

Note: To use PuppetDB, the nodes at your site must be running Puppet 3.5.1 or later.

PuppetDB can also be used with standalone Puppet deployments where each node runs puppet apply. Once connected to PuppetDB, puppet apply will do the following:

  • Send the node's catalog to PuppetDB
  • Query PuppetDB when compiling catalogs that collect exported resources
  • Store facts in PuppetDB
  • Send reports to PuppetDB (optional)

You will need to take the following steps to configure your standalone nodes to connect to PuppetDB. Note that since you must change Puppet's configuration on every managed node, we strongly recommend that you do so with Puppet itself.

Step 1: Configure SSL

PuppetDB requires client authentication for its SSL connections and the PuppetDB termini require SSL to talk to PuppetDB. You must configure Puppet and PuppetDB to work around this double-bind by using one of the following options:

Option A: Set Up an SSL Proxy for PuppetDB

  1. Edit the jetty section of the PuppetDB config files to remove all SSL-related settings.
  2. Install a general purpose web server (like Apache or Nginx) on the PuppetDB server.
  3. Configure the web server to listen on port 8081 with SSL enabled and proxy all traffic to localhost:8080 (or whatever unencrypted hostname and port were set in jetty.ini). The proxy server can use any certificate --- as long as Puppet has never downloaded a CA cert from a puppet master, it will not verify the proxy server's certificate. If your nodes have downloaded CA certs, you must either make sure the proxy server's cert was signed by the same CA, or delete the CA cert.

Option B: Issue Certificates to All Puppet Nodes

When talking to PuppetDB, puppet apply can use the certificates issued by a puppet master's certificate authority . You can issue certificates to every node by setting up a puppet master server with dummy manifests, running puppet agent --test once on every node, signing every certificate request on the puppet master, and running puppet agent --test again on every node.

Do the same on your PuppetDB node, then re-run the SSL setup script. PuppetDB will now trust connections from your Puppet nodes.

You will have to sign a certificate for every new node you add to your site.

Step 2: Install Terminus Plugins on Every Puppet Node

Currently, Puppet needs extra Ruby plugins in order to use PuppetDB. Unlike custom facts or functions, these cannot be loaded from a module and must be installed in Puppet's main source directory.

    package {'puppetdb-termini':
      ensure => installed,
    }

On Platforms Without Packages

If your puppet master isn't running Puppet from a supported package, you will need to install the plugins using file resources.

  • Download the PuppetDB source code; unzip it, locate the puppet/lib/puppet directory and put it in the files directory of the Puppet module you are using to enable PuppetDB integration.
  • Identify the install location of Puppet on your nodes.
  • Create a file resource in your manifests for each of the plugin files, to move them into place on each node.
    # <modulepath>/puppetdb/manifests/terminus.pp
    class puppetdb::terminus {
      $puppetdir = "$rubysitedir/puppet"

      file { $puppetdir:
        ensure => directory,
        recurse => remote, # Copy these files without deleting the existing files
        source => "puppet:///modules/puppetdb/puppet",
        owner => root,
        group => root,
        mode => 0644,
      }
    }

Step 3: Manage Config Files on Every Puppet Node

All of the config files you need to manage will be in Puppet's config directory (confdir). When managing these files with puppet apply, you can use the $settings::confdir variable to automatically discover the location of this directory.

Manage puppetdb.conf

You can specify the contents of puppetdb.conf directly in your manifests. It should contain the PuppetDB server's hostname and port:

[main]
server = puppetdb.example.com
port = 8081

PuppetDB's port for secure traffic defaults to 8081. Puppet requires use of PuppetDB's secure, HTTPS port. You cannot use the unencrypted, plain HTTP port.

For availability reasons there is a setting named soft_write_failure that will cause the PuppetDB termini to fail in a soft-manner if PuppetDB is not accessable for command submission. This will mean that users who are either not using storeconfigs, or only exporting resources will still have their catalogs compile during a PuppetDB outage.

If no puppetdb.conf file exists, the following default values will be used:

server = puppetdb
port = 8081
soft_write_failure = false

Manage puppet.conf

You will need to create a template for puppet.conf based on your existing configuration. Then, modify the template by adding the following settings to the [main] block:

[main]
  storeconfigs = true
  storeconfigs_backend = puppetdb
  # Optional settings to submit reports to PuppetDB:
  report = true
  reports = puppetdb

Note: The thin_storeconfigs and async_storeconfigs settings should be absent or set to false.

Manage routes.yaml

Typically, you can specify the contents of routes.yaml directly in your manifests; if you are already using it for some other purpose, you will need to manage it with a template based on your existing configuration. The path to this Puppet configuration file can be found with the command puppet master --configprint route_file.

Ensure that the following keys are present:

---
apply:
  catalog:
    terminus: compiler
    cache: puppetdb
  resource:
    terminus: ral
    cache: puppetdb
  facts:
    terminus: facter
    cache: puppetdb_apply

This is necessary to keep Puppet from using stale facts and to keep the puppet resource subcommand from malfunctioning. Note that the puppetdb_apply terminus is specifically for puppet apply nodes, and differs from the configuration of puppet masters using PuppetDB.