This module provides a collection of reusable profiles for common applications.
For most of the profiles is not needed a dedicated component module.
Prerequites for this module are example42's tp and psick modules.
This module manages plenty of different profiles for many common applications. You can cherry pick which ones to use and make them cohexist with your own profiles or with component modules.
For documentation on the specific application profiles refere to the relevant docs pages or directly in the code comments, when present.
Every psick profile can be classified and used indipendently.
You need to classify also the psick class from the psick module (which by default, without specific Hiera data it does nothing) in order to leverage on the general variables set in the main psick class and used in the psick modules.
In short in you manifests/site.pp or wherever you classify your nodes you need to:
include psick
and then include/classify the psick profiles you want, for example:
include psick_profile::gitlab
include psick_profile::grafana
Every psick profile manages the relevant application.
In some cases you have the option to decide if to use an external component module to install it or use Tiny Puppet.
Refer to each profile documentation for more info on the managed resources.
The classes when an application is installed using Tiny Puppet are always called tp and are placed in manifests called tp.pp
For example the class psick_profile::jenkins::tp, defined under manifests/jenkins/tp.pp manages with installation of Jenkins via Tiny Puppet and is, by defauly, included from the main psick_profile::jenkins class.
Psick_profile module requires:
- example42-psick module
- example42-tp module (which needs the example42-tinydata module)
The above, of course, need stdlib, which you probably are already using:
- puppetlabs/stdlib
According to the OS used you might need other modules:
- puppetlabs/vcsrepo (if tp::dir define is used with vcsrepos)
- puppetlabs/concat (might be needed in some profiles)
- puppetlabs/chocolatey (on Windows nodes)
- homebrew module (on Darwin nodes)
Some profiles might require an additional component modules.
Refer to Puppetfile in the docs dir for the complete reference of needed modules, in Puppetfile format.
Use whatever classification approach you want and classify the psick profile you want to use.
Remember to classify the psick class as well.
Considering that psick class can be used also for classification, all you might need is something as follows:
In your control-repos' manifests/site.pp just classify psick for all nodes:
node default{
include psick
}
An then manage everything via Hiera (refer to psick documentation for details), classification included, with something like:
# Psick based classification for Linux nodes:
psick::pre::linux_classes:
puppet: psick::puppet
hostname: psick::hostname
hosts: psick::hosts::resource
dns: psick::dns::resolver
repo: psick::repo
users: psick::users
psick::base::linux_classes:
ssh: psick::openssh
sudo: psick::sudo
time: psick::time
sysctl: psick::sysctl
update: psick::update
motd: psick::motd
selinux: psick::selinux
limits: psick::limits
systat: psick_profile::sar
mail: psick_profile::postfix
icinga: psick_profile::icinga2
monitor_plugins: psick_profile::nagiosplugins
tp: tp
# Psick based classification for Windows nodes
psick::pre::windows_classes:
hosts: psick::hosts::resource
chocolatey: chocolatey
psick::base::windows_classes:
features: psick::windows::features
registry: psick::windows::registry
services: psick::windows::services
tp: tp
# Psick based classification for MacOS nodes
psick::pre::darwin_classes:
homebrew: homebrew
puppet: psick::puppet
psick::base::darwin_classes:
tp: tp
Note that each of the above Hiera keys (looked up in Deep merge mode) allows you to classify classes for different OSes (Linux, Windows, MacOS) in different stages, applied in order (pre, base, profile).
The value of each Hiera key is an hash of key values: the keys can be any string and you can use to override the classes to include at different Hiera levels. The values are simply the classes to classify: they can be your own profiles, a componenent module class, a profile from the psick module or a profile from this module.
Look at the single profiles in code documentation or at the directory in docs.
Not all the profiles are tested or work on every OS supported by this module.
To contribute to this module open a Pull Request on GitHub, and follow the instructions there.