Don't Panic — Puppet Enterprise

Puppet Enterprise, usually shortened to PE, is the part of Puppet that turns a collection of configuration files into an operated system. Puppet code still describes the state a node should have. The important addition is the machinery around it: deployment, classification, evidence, access control, and enough moving parts to make a small orchestra manager look modest.

The central idea is a catalog. It is the node-specific set of resources Puppet compiles from code, data, facts, and classification. An agent applies that catalog and sends back a report. This is why PE is not a different Puppet language. It is the governed route from a Git change to a recorded result.

Start with the control repository, the Git repository that holds environments, Hiera data, and the Puppetfile. Code Manager turns a reviewed branch into a deployed environment. File sync distributes that environment. Compilers can then produce catalogs for agents. Editing a managed environment directory by hand has the durability of writing a note on a departing train: the next deployment can replace it.

The surprise is that compilation is not success. A catalog can be perfectly valid and an agent can still fail later, or repair drift that somebody created outside policy. PuppetDB, the service that stores facts, catalogs, and reports, supplies the evidence. Resource events and corrective changes tell you what nodes did, not merely what the control plane hoped they would do.

PE also separates continuing state from deliberate action. Scheduled agents keep modeled state true. The orchestrator runs Puppet jobs, tasks, and plans when timing, order, or explicit control matters. Add RBAC, role-based access control, and the system can distinguish who may deploy code, classify nodes, start jobs, or only inspect results. That distinction is not decoration; it is the boundary between an operated platform and a shared administrator password wearing a nicer hat.

The rest of the course fills in the map. Read the intro for the primary server, compilers, PuppetDB, recovery, and scale path. Use the slides to follow the delivery loop at a glance. Keep the cheatsheet nearby when separating node groups, health signals, and orchestration choices. Field Notes focuses on the failure patterns that make a seemingly tidy platform behave less tidily in production.