9. Roles and Profiles
by Thomas Uphillis, Jaime Soriano Pastor, Alessandro Franceschi, Martin Alfke, Feli
Puppet: Mastering Infrastructure Automation
- Puppet: Mastering Infrastructure Automation
- Table of Contents
- Puppet: Mastering Infrastructure Automation
- Puppet: Mastering Infrastructure Automation
- Credits
- Preface
- 1. Module 1
- 1. Writing Your First Manifests
- 2. The Master and Its Agents
- 3. A Peek under the Hood – Facts, Types, and Providers
- 4. Modularizing Manifests with Classes and Defined Types
- 5. Extending Your Puppet Infrastructure with Modules
- An overview of Puppet's modules
- Maintaining environments
- Building a specific module
- Naming your module
- Making your module available to Puppet
- Implementing basic module functionality
- Creating utilities for derived manifests
- Enhancing the agent through plugins
- Enhancing Puppet's system knowledge through facts
- Refining the interface of your module through custom functions
- Making your module portable across platforms
- Finding helpful Forge modules
- Summary
- 6. Leveraging the Full Toolset of the Language
- 7. New Features from Puppet 4
- 8. Separating Data from Code Using Hiera
- 2. Module 2
- 1. Puppet Essentials
- The Puppet ecosystem
- Puppet components
- Installing and configuring Puppet
- Puppet in action
- Variables, facts, and scopes
- Meta parameters
- Managing order and dependencies
- Reserved names and allowed characters
- Conditionals
- Comparison operators
- Iteration and lambdas
- Exported resources
- Modules
- Restoring files from a filebucket
- Summary
- 2. Managing Puppet Data with Hiera
- 3. Introducing PuppetDB
- 4. Designing Puppet Architectures
- 5. Using and Writing Reusable Modules
- 6. Higher Abstraction Modules
- 7. Puppet Migration Patterns
- 8. Code Workflow Management
- 9. Scaling Puppet Infrastructures
- 10. Extending Puppet
- 11. Beyond the System
- 12. Future Puppet
- 1. Puppet Essentials
- 3. Module 3
- 1. Dealing with Load/Scale
- 2. Organizing Your Nodes and Data
- 3. Git and Environments
- 4. Public Modules
- 5. Custom Facts and Modules
- 6. Custom Types
- 7. Reporting and Orchestration
- 8. Exported Resources
- 9. Roles and Profiles
- 10. Troubleshooting
- A. Bibliography
- Index
In Chapter 2, Organizing Your Nodes and Data, we showed you how to organize your nodes using an ENC or Hiera, or ideally both. At that point, we didn't cover the Forge modules or writing your own modules, as we did in Chapter 4, Public Modules, and Chapter 5, Custom Facts and Modules. In this chapter, we will cover a popular design concept employed in large installations of Puppet. The idea was originally made popular by Craig Dunn in his blog, which can be found at http://www.craigdunn.org/2012/05/239/. Garry Larizza also wrote a useful post on the subject at http://garylarizza.com/blog/2014/02/17/puppet-workflow-part-2/.
The concept put forth by Craig Dunn in his blog is the one at which most Puppet masters arrive independently. Modules should be nested in such a way that common components can be shared among nodes. The naming convention that is generally accepted is that roles contain one or more profiles. Profiles in turn contain one or more modules. You can have a node-level logic that is very clean and elegant using the roles and profile design patterns, together with an ENC and Hiera,. The ENC and/or Hiera can also be used to enforce standards on your nodes without interfering with the roles and profiles. As we discussed in Chapter 2, Organizing Your Nodes and Data, with the virtual module it is possible to have Hiera apply classes automatically to any system where the is_virtual fact is true. Applying the same logic to facts such as osfamily, we can ensure that all the nodes for which osfamily is RedHat, receive an appropriate module.
Putting all these elements together, we arrive at the following diagram showing how modules are applied to a node:
Roles are the high-level abstraction of what a node will do.
-
No Comment