![]() master bcbe7aa35c03c83dd58842fdb0f72a332e660124 source: https://github.com/stackforge/puppet-openstacklib Partially Implements: blueprint upgrade-openstack-puppet-modules Change-Id: I6eea00337c5cce2269c236b49305ba7085f68eb4 |
||
---|---|---|
.. | ||
lib/puppet | ||
manifests | ||
spec | ||
.fixtures.yml | ||
.gitignore | ||
Gemfile | ||
LICENSE | ||
metadata.json | ||
Rakefile | ||
README.md |
openstacklib
5.1.0 - 2014.2 - Juno
Table of Contents
- Overview - What is the openstacklib module?
- Module Description - What does the module do?
- Setup - The basics of getting started with openstacklib
- Implementation - An under-the-hood peek at what the module is doing
- Limitations - OS compatibility, etc.
- Development - Guide for contributing to the module
- Contributors - Those with commits
- Release Notes - Notes on the most recent updates to the module
Overview
The openstacklib module is a part of Stackforge, an effort by the Openstack infrastructure team to provide continuous integration testing and code review for Openstack and Openstack community projects not part of the core software. The module itself is used to expose common functionality between Openstack modules as a library that can be utilized to avoid code duplication.
Module Description
The openstacklib module is a library module for other Openstack modules to utilize. A thorough description will be added later.
This module is tested in combination with other modules needed to build and leverage an entire Openstack software stack. These modules can be found, all pulled together in the openstack module.
Setup
Installing openstacklib
example% puppet module install puppetlabs/openstacklib
Usage
Classes and Defined Types
Defined type: openstacklib::db::mysql
The db::mysql resource is a library resource that can be used by nova, cinder, ceilometer, etc., to create a mysql database with configurable privileges for a user connecting from defined hosts.
Typically this resource will be declared with a notify parameter to configure the sync command to execute when the database resource is changed.
For example, in heat::db::mysql you might declare:
::openstacklib::db::mysql { 'heat':
password_hash => mysql_password($password),
dbname => $dbname,
user => $user,
host => $host,
charset => $charset,
collate => $collate,
allowed_hosts => $allowed_hosts,
notify => Exec['heat-dbsync'],
}
Some modules should ensure that the database is created before the service is set up. For example, in keystone::db::mysql you would have:
::openstacklib::db::mysql { 'keystone':
password_hash => mysql_password($password),
dbname => $dbname,
user => $user,
host => $host,
charset => $charset,
collate => $collate,
allowed_hosts => $allowed_hosts,
notify => Exec['keystone-manage db_sync'],
before => Service['keystone'],
}
** Parameters for openstacklib::db::mysql: **
#####password_hash
Password hash to use for the database user for this service;
string; required
#####dbname
The name of the database
string; optional; default to the $title of the resource, i.e. 'nova'
#####user
The database user to create;
string; optional; default to the $title of the resource, i.e. 'nova'
#####host
The IP address or hostname of the user in mysql_grant;
string; optional; default to '127.0.0.1'
#####charset
The charset to use for the database;
string; optional; default to 'utf8'
#####collate
The collate to use for the database;
string; optional; default to 'utf8_general_ci'
#####allowed_hosts
Additional hosts that are allowed to access this database;
array or string; optional; default to undef
#####privileges
Privileges given to the database user;
string or array of strings; optional; default to 'ALL'
Defined type: openstacklib::db::postgresql
The db::postgresql resource is a library resource that can be used by nova, cinder, ceilometer, etc., to create a postgresql database and a user with configurable privileges.
Typically this resource will be declared with a notify parameter to configure the sync command to execute when the database resource is changed.
For example, in heat::db::postgresql you might declare:
::openstacklib::db::postgresql { $dbname:
password_hash => postgresql_password($user, $password),
dbname => $dbname,
user => $user,
notify => Exec['heat-dbsync'],
}
Some modules should ensure that the database is created before the service is set up. For example, in keystone::db::postgresql you would have:
::openstacklib::db::postgresql { $dbname:
password_hash => postgresql_password($user, $password),
dbname => $dbname,
user => $user,
notify => Exec['keystone-manage db_sync'],
before => Service['keystone'],
}
** Parameters for openstacklib::db::postgresql: **
#####password_hash
Password hash to use for the database user for this service;
string; required
#####dbname
The name of the database
string; optional; default to the $title of the resource, i.e. 'nova'
#####user
The database user to create;
string; optional; default to the $title of the resource, i.e. 'nova'
#####encoding
The encoding use for the database;
string; optional; default to undef
#####privileges
Privileges given to the database user;
string or array of strings; optional; default to 'ALL'
Defined type: openstacklib::service_validation
The service_validation resource is a library resource that can be used by nova, cinder, ceilometer, etc., to validate that a resource is actually up and running.
For example, in nova::api you might declare:
::openstacklib::service_validation { 'nova-api':
command => 'nova list',
}
This defined resource creates an exec-anchor pair where the anchor depends upon the successful exec run.
** Parameters for openstacklib::service_validation: **
#####command
Command to run for validating the service;
string; required
#####service_name
The name of the service to validate;
string; optional; default to the $title of the resource, i.e. 'nova-api'
#####path
The path of the command to validate the service;
string; optional; default to '/usr/bin:/bin:/usr/sbin:/sbin'
#####provider
The provider to use for the exec command;
string; optional; default to 'shell'
#####tries
Number of times to retry validation;
string; optional; default to '10'
#####try_sleep
Number of seconds between validation attempts;
string; optional; default to '2'
Types and Providers
Aviator
#####Puppet::add_aviator_params
The aviator type is not a real type, but it serves to simulate a mixin model, whereby other types can call out to the Puppet::add_aviator_params method in order to add aviator-specific parameters to themselves. Currently this adds the auth parameter to the given type. The method must be called after the type is declared, e.g.:
require 'puppet/type/aviator'
Puppet::Type.newtype(:my_type) do
# ...
end
Puppet::add_aviator_params(:my_type)
#####Puppet::Provider::Aviator
The aviator provider is a parent provider intended to serve as a base for other providers that need to authenticate against keystone in order to accomplish a task.
Puppet::Provider::Aviator#authenticate
Either creates an authenticated session or sets up an unauthenticated session with instance variables initialized with a token to inject into the next request. It takes as arguments a set of authentication parameters as a hash and a path to a log file. Puppet::Provider::Aviator#authencate looks for five different possible methods of authenticating, in the following order:
- Username and password credentials in the auth parameters
- The path to an openrc file containing credentials to read in the auth parameters
- A service token in the auth parameters
- Environment variables set for the environment in which Puppet is running
- A service token in /etc/keystone/keystone.conf. This option provides backwards compatibility with earlier keystone providers.
If the provider has password credentials, it can create an authenticated session. If it only has a service token, it initializes an unauthenciated session and a hash of session data that can be injected into a future request.
Puppet::Provider::Aviator#make_request
After creating a session, the make_request method provides an interface that providers can use to make requests without worrying about whether they have an authenticated or unauthenticated session. It takes as arguments the Aviator::Service it is making a request at (for example, keystone), a symbol for the request (for example, :list_tenants), and optionally a block to execute that will set parameters for an update request.
Implementation
openstacklib
openstacklib is a combination of Puppet manifest and ruby code to delivery configuration and extra functionality through types and providers.
Limitations
The python-migrate system package for RHEL 6 and below is out of date and may fail to correctly migrate postgresql databases. While this module does not handle database migrations, it is common to set up refresh relationships between openstacklib::db::postgresql resource and the database sync exec resource. Relying on this behavior may cause errors.
Beaker-Rspec
This module has beaker-rspec tests
To run:
bundle install
bundle exec rspec spec/acceptance
Development
Developer documentation for the entire puppet-openstack project.
Contributors
Versioning
This module has been given version 5 to track the puppet-openstack modules. The versioning for the puppet-openstack modules are as follows:
Puppet Module :: OpenStack Version :: OpenStack Codename
2.0.0 -> 2013.1.0 -> Grizzly
3.0.0 -> 2013.2.0 -> Havana
4.0.0 -> 2014.1.0 -> Icehouse
5.0.0 -> 2014.2.0 -> Juno
Release Notes
5.1.0
- Update .gitreview file for project rename
- Adding augeas insertion check
- MySQL: change default MySQL collate to utf8_general_ci
- spec: pin rspec-puppet to 1.0.1
5.0.0
- This is the initial release of this module.