Where does a Puppet agent run start from?

A Puppet agent run starts on the node being managed, which is the correct answer. Each managed node runs its own Puppet agent service that initiates the agent run process independently. This process involves the agent periodically contacting the Puppet master to retrieve the latest configuration settings and resources defined in Puppet manifests.

Once the agent retrieves the necessary information from the master, it applies the configurations to ensure that the node aligns with the desired state as defined in the Puppet code. This agent-centric approach allows for decentralized management, providing flexibility in how and when configuration updates are applied across different nodes in the infrastructure.

In contrast, the master server acts as a central point of configuration but does not initiate the agent runs. PuppetDB serves as a database for storing and querying Puppet data, while the console generally refers to the user interface for managing Puppet configurations and data, neither of which directly execute the agent runs. Thus, the agent run starts at the managed node itself.