Implied Labels

Infer dynamic node labels using configured implications.

Lets have a node with linux, RHEL, RHEL8 and RHEL8_5 labels. Most of this information is redundant and can be inferred from the RHEL8_5 label. Here is an excerpt of the Implied Labels Plugin configuration to do just that.

This plugin gathers the implication rules to one place (Manage Jenkins > Label implications), so explicit node labels can be defined in more concise form (without redundancy). Implications are readable for every user with Jenkins.READ, but only Jenkins.ADMINISTER is authorized to configure. The plugin also detects explicitly configured node labels that can be inferred using existing implication rules.

Details

Jenkins administrators can declare any number of implication rules for label inference. An implication consists of label expression and atom list. A node that matches label expression will have assigned new labels from atom list. Labels contributed from the Implied Labels Plugin are dynamic labels and thus not saved in configuration. Implications are evaluated in topological order so implication expressions can refer to labels contributed by other implications.

Configuration as code

Label definitions can be automated with configuration as code. The platform labeler can automatically assign labels based on operating system properties and those operating system properties can be used to define more labels.

unclassified:
  impliedLabels:
    implications:
    - atoms: "linux"
      expression: "Ubuntu || CentOS || Debian || Rocky || openSUSE"