Puppet is still one of the most widely deployed configuration management tools in large enterprises, especially in banking, telecom and government estates across the US and UK. Interviewers expect you to understand both the language and how Puppet runs at scale. The questions below are grouped from fundamentals to advanced, with short answers you can expand on using your own project experience.
Key takeaways
- Puppet is declarative: you describe the desired state and the agent makes the system match it.
- Know the agent and server flow, catalogs, facts, Hiera and PuppetDB well; they come up in almost every interview.
- Puppet has been owned by Perforce since 2022. In 2025 Puppet moved new builds to a private, controlled location, and community members started an open source fork called OpenVox.
- Be ready to compare Puppet with Ansible, Chef and Terraform and explain where each fits.
Fundamentals
1. What is Puppet?
Puppet is a configuration management tool that uses a declarative domain-specific language to define the desired state of servers. An agent on each node applies that state and reports back, so drift is corrected automatically.
2. What is a manifest?
A manifest is a file with a .pp extension that contains Puppet code: resources, classes and defined types. The main manifest, usually site.pp, is the entry point that assigns classes to nodes.
3. What is a resource?
A resource is the basic unit of configuration, such as a package, file, service or user. Each has a type, a title and attributes, for example a package resource with ensure set to installed.
4. What is a module?
A module is a directory with a standard layout (manifests, files, templates, lib, data) that packages reusable code for one component. Modules are shared on the Puppet Forge.
5. What are classes and defined types?
A class is a named block of code applied once per node. A defined type is a reusable block that can be declared many times with different titles, such as several virtual hosts.
Architecture and workflow
6. Explain the Puppet agent run.
The agent gathers facts with Facter and sends them to the Puppet server over HTTPS (port 8140 by default). The server compiles a catalog for that node, the agent applies it, and a report is sent back. By default the agent runs every 30 minutes.
7. What is a catalog?
A catalog is the compiled, node-specific document listing every resource and the relationships between them. The agent applies the catalog, not the raw code.
8. What is Facter?
Facter collects system information such as OS, IP addresses and memory. Facts are available in code as variables and through the facts hash, and you can write custom or external facts.
9. What is Hiera and why use it?
Hiera is Puppet’s hierarchical key value lookup system. It separates data from code so the same module can behave differently per environment, data center or node. Hiera 5 supports a layered hierarchy at global, environment and module level.
10. What is PuppetDB?
PuppetDB stores facts, catalogs and reports. It enables exported resources and lets you query infrastructure data, for example to build monitoring configuration from every node that runs a web service.
11. How are certificates handled?
On first contact the agent creates a certificate signing request. The server’s certificate authority signs it, manually or through autosigning rules, and all later traffic uses mutual TLS.
Intermediate and advanced
12. How do you control resource ordering?
Use the metaparameters require, before, notify and subscribe, or chaining arrows. Notify and subscribe also trigger refreshes, for example restarting a service when its config file changes.
13. What are environments?
Environments are isolated sets of code and data, usually mapped to Git branches. Tools such as r10k or Code Manager in Puppet Enterprise deploy each branch as an environment.
14. What is Bolt?
Bolt is Puppet’s agentless task runner. It runs ad hoc commands, scripts, tasks and plans over SSH or WinRM, which helps with one-time jobs that do not fit the desired-state model.
15. How do you test Puppet code?
Use puppet parser validate and puppet-lint for syntax and style, rspec-puppet for unit tests, and acceptance testing with Litmus. The Puppet Development Kit (PDK) scaffolds modules with these tools.
16. What is the difference between Puppet and Ansible?
Puppet is agent based, declarative and model driven, with continuous drift correction. Ansible is agentless, uses YAML playbooks over SSH and is often run on demand. Puppet suits large fleets that must stay compliant; Ansible suits orchestration and quick automation.
17. How does Puppet compare to Terraform?
Terraform provisions infrastructure such as networks and virtual machines through cloud APIs. Puppet configures what runs inside those machines. Many teams use both together.
18. What is OpenVox?
OpenVox is a community fork of open source Puppet, started in 2025 after Puppet’s owner began shipping new binaries and packages from a private, controlled location. It aims to keep open source Puppet development going and stays compatible with existing Puppet code.
19. Describe a real troubleshooting scenario.
A strong answer walks through running puppet agent -t with debug output, checking the report for failed resources, confirming facts and Hiera lookups with puppet lookup, and fixing ordering or dependency issues.
How to prepare
Build a small lab with one server and two agents, write a module with Hiera data, and practice an r10k deployment. Pair Puppet with cloud and container skills from our Terraform training, Kubernetes training and Azure DevOps training. All are live and instructor-led with hands-on labs, certification guidance and a free demo. You can browse more guides on the Tutorials Mania blog.