Librarian-puppet

Introduction

Librarian-puppet is a bundler for your puppet infrastructure. You can use librarian-puppet to manage the puppet modules your infrastructure depends on, whether the modules come from the Puppet Forge, Git repositories or just a path.

Librarian-puppet can reuse the dependencies listed in your Modulefile or metadata.json
Forge modules can be installed from Puppetlabs Forge or an internal Forge such as Pulp
Git modules can be installed from a branch, tag or specific commit, optionally using a path inside the repository
Modules can be installed from GitHub using tarballs, without needing Git installed
Modules can be installed from a filesystem path
Module dependencies are resolved transitively without needing to list all the modules explicitly

Librarian-puppet manages your modules/ directory for you based on your Puppetfile. Your Puppetfile becomes the authoritative source for what modules you require and at what version, tag or branch.

Once using Librarian-puppet you should not modify the contents of your modules directory. The individual modules' repos should be updated, tagged with a new release and the version bumped in your Puppetfile.

It is based on Librarian, a framework for writing bundlers, which are tools that resolve, fetch, install, and isolate a project's dependencies.

Versions

Librarian-puppet >= 2.0 (as well as 1.1, 1.2 and 1.3) requires Ruby 1.9 and uses the Puppet Forge API v3. Versions < 2.0 work on Ruby 1.8.

See the Changelog for more details.

The Puppetfile

Every Puppet repository that uses Librarian-puppet may have a file named Puppetfile, metadata.json or Modulefile in the root directory of that repository. The full specification for which modules your puppet infrastructure repository depends goes in here.

Simple usage

If no Puppetfile is present, librarian-puppet will download all the dependencies listed in your metadata.json or Modulefile from the Puppet Forge, as if the Puppetfile contained

forge "https://forgeapi.puppetlabs.com"

metadata

Example Puppetfile

forge "https://forgeapi.puppetlabs.com"

mod 'puppetlabs-razor'
mod 'puppetlabs-ntp', "0.0.3"

mod 'puppetlabs-apt',
  :git => "git://github.com/puppetlabs/puppetlabs-apt.git"

mod 'puppetlabs-stdlib',
  :git => "git://github.com/puppetlabs/puppetlabs-stdlib.git"

mod 'puppetlabs-apache', '0.6.0',
  :github_tarball => 'puppetlabs/puppetlabs-apache'

mod 'acme-mymodule', :path => './some_folder'

exclusion 'acme-bad_module'

Recursive module dependency resolution

When fetching a module all dependencies specified in its Modulefile, metadata.json and Puppetfile will be resolved and installed.

Puppetfile Breakdown

forge "https://forgeapi.puppetlabs.com"

This declares that we want to use the official Puppet Labs Forge as our default source when pulling down modules. If you run your own local forge, you may want to change this.

metadata

Download all the dependencies listed in your metadata.json or Modulefile from the Puppet Forge.

mod 'puppetlabs-razor'

Pull in the latest version of the Puppet Labs Razor module from the default source.

mod 'puppetlabs-ntp', "0.0.3"

Pull in version 0.0.3 of the Puppet Labs NTP module from the default source.

mod 'puppetlabs-apt',
  :git => "git://github.com/puppetlabs/puppetlabs-apt.git"

Our puppet infrastructure repository depends on the apt module from the Puppet Labs GitHub repos and checks out the master branch.

mod 'puppetlabs-apt',
  :git => "git://github.com/puppetlabs/puppetlabs-apt.git",
  :ref => '0.0.3'

Our puppet infrastructure repository depends on the apt module from the Puppet Labs GitHub repos and checks out a tag of 0.0.3.

mod 'puppetlabs-apt',
  :git => "git://github.com/puppetlabs/puppetlabs-apt.git",
  :ref => 'feature/master/dans_refactor'

Our puppet infrastructure repository depends on the apt module from the Puppet Labs GitHub repos and checks out the dans_refactor branch.

When using a Git source, we do not have to use a :ref =>. If we do not, then librarian-puppet will assume we meant the master branch.

If we use a :ref =>, we can use anything that Git will recognize as a ref. This includes any branch name, tag name, SHA, or SHA unique prefix. If we use a branch, we can later ask Librarian-puppet to update the module by fetching the most recent version of the module from that same branch.

Note that Librarian-puppet recognizes the r10k Puppetfile's additional options, :tag, :commit, and :branch, but only as aliases for :ref. That is, there is no implementation of r10k's optimizations around fetching these different types of git objects.

The Git source also supports a :path => option. If we use the path option, Librarian-puppet will navigate down into the Git repository and only use the specified subdirectory. Some people have the habit of having a single repository with many modules in it. If we need a module from such a repository, we can use the :path => option here to help Librarian-puppet drill down and find the module subdirectory.

mod 'puppetlabs-apt',
  :git => "git://github.com/fake/puppet-modules.git",
  :path => "modules/apt"

Our puppet infrastructure repository depends on the apt module, which we have stored as a directory under our puppet-modules git repos.

mod 'puppetlabs-apache', '0.6.0',
  :github_tarball => 'puppetlabs/puppetlabs-apache'

Our puppet infrastructure repository depends on the puppetlabs-apache module, to be downloaded from GitHub tarball.

mod 'acme-mymodule', :path => './some_folder'

Our puppet infrastructure repository depends on the acme-mymodule module, which is already in the filesystem.

exclusion 'acme-bad_module'

Exclude the module acme-bad_module from resolution and installation.

How to Use

Install librarian-puppet:

$ gem install librarian-puppet

Prepare your puppet infrastructure repository:

$ cd ~/path/to/puppet-inf-repos
$ (git) rm -rf modules
$ librarian-puppet init

Librarian-puppet takes over your modules/ directory, and will always reinstall (if missing) the modules listed the Puppetfile.lock into your modules/ directory, therefore you do not need your modules/ directory to be tracked in Git.

Librarian-puppet uses a .tmp/ directory for tempfiles and caches. You should not track this directory in Git.

Running librarian-puppet init will create a skeleton Puppetfile for you as well as adding tmp/ and modules/ to your .gitignore.

$ librarian-puppet install [--clean] [--verbose]

This command looks at each mod declaration and fetches the module from the source specified. This command writes the complete resolution into Puppetfile.lock and then copies all of the fetched modules into your modules/ directory, overwriting whatever was there before.

Librarian-puppet support both v1 and v3 of the Puppet Forge API. Specify a specific API version when installing modules:

$ librarian-puppet install --use-v1-api # this is default; ignored for official Puppet Forge
$ librarian-puppet install --no-use-v1-api # use the v3 API; default for official Puppet Forge

Please note that this does not apply for the official Puppet Forge, where v3 is used by default.

Get an overview of your Puppetfile.lock with:

$ librarian-puppet show

Inspect the details of specific resolved dependencies with:

$ librarian-puppet show NAME1 [NAME2, ...]

Find out which dependencies are outdated and may be updated:

$ librarian-puppet outdated [--verbose]

Update the version of a dependency:

$ librarian-puppet update apt [--verbose]
$ git diff Puppetfile.lock
$ git add Puppetfile.lock
$ git commit -m "bumped the version of apt up to 0.0.4."

Configuration

Configuration comes from three sources with the following highest-to-lowest precedence:

The local config (./.librarian/puppet/config)
The environment
The global config (~/.librarian/puppet/config)

You can inspect the final configuration with:

$ librarian-puppet config

You can find out where a particular key is set with:

$ librarian-puppet config KEY

You can set a key at the global level with:

$ librarian-puppet config KEY VALUE --global

And remove it with:

$ librarian-puppet config KEY --global --delete

You can set a key at the local level with:

$ librarian-puppet config KEY VALUE --local

And remove it with:

$ librarian-puppet config KEY --local --delete

You cannot set or delete environment-level config keys with the CLI.

Configuration set at either the global or local level will affect subsequent invocations of librarian-puppet. Configurations set at the environment level are not saved and will not affect subsequent invocations of librarian-puppet.

You can pass a config at the environment level by taking the original config key and transforming it: replace hyphens (-) with underscores (_) and periods (.) with doubled underscores (__), uppercase, and finally prefix with LIBRARIAN_PUPPET_. For example, to pass a config in the environment for the key part-one.part-two, set the environment variable LIBRARIAN_PUPPET_PART_ONE__PART_TWO.

Configuration affects how various commands operate.

The path config sets the directory to install to. If a relative path, it is relative to the directory containing the Puppetfile. The equivalent environment variable is LIBRARIAN_PUPPET_PATH.
The tmp config sets the cache directory for librarian. If a relative path, it is relative to the directory containing the Puppetfile. The equivalent environment variable is LIBRARIAN_PUPPET_TMP.

Configuration can be set by passing specific options to other commands.

The path config can be set at the local level by passing the --path option to the install command. It can be unset at the local level by passing the --no-path option to the install command. Note that if this is set at the environment or global level then, even if --no-path is given as an option, the environment or global config will be used.

Rsync Option

The default convergence strategy between the cache and the module directory is to execute an rm -r on the module directory and just cp -r from the cache. This causes the module to be removed from the module path every time librarian puppet updates, regardless of whether the content has changed. This can cause some problems in environments with lots of change. The problem arises when the module directory gets removed while Puppet is trying to read files inside it. The puppet master process will lose its CWD and the catalog will fail to compile. To avoid this, you can use rsync to implement a more conservative convergence strategy. This will use rsync with the -avz and --delete flags instead of a rm -r and cp -r. To use this feature, just set the rsync configuration setting to true.

$ librarian-puppet config rsync true --global

Alternatively, using an environment variable:

LIBRARIAN_PUPPET_RSYNC='true'

Note that the directories will still be purged if you run librarian-puppet with the --clean or --destructive flags.

How to Contribute

Pull requests please.
Bonus points for feature branches.

Reporting Issues

Bug reports to the github issue tracker please. Please include:

Relevant Puppetfile and Puppetfile.lock files
Version of ruby, librarian-puppet, and puppet
What distro
Please run the librarian-puppet commands in verbose mode by using the --verbose flag, and include the verbose output in the bug report as well.

License

Please see the LICENSE file.

brianhks / librarian-puppet