Aggregating Pipeline Configuration¶
Build the Aggregated Pipeline Configuration¶
PipelineConfigurationAggregator handles fetching the configuration hierarchy from Governance Tiers and the job itself.
For example, a given job would get its parent chain all the way up to the Jenkins instance's Global Configuration.
flowchart TD
g[Global Configuration]
g --> f_a[Folder A]
g --> f_b[Folder B]
f_a --> j_a[Job A]
f_a --> j_b[Job B]
f_b --> j_c[JobC]
f_b --> j_d[Job D]
style j_b fill:#FFD580,stroke:#000000Turns into:
flowchart LR
g[Global Configuration]
g --> f_a[Folder A]
f_a --> j_b[Job B]
style j_b fill:#FFD580,stroke:#000000If a Governance Tier or the job itself have a Pipeline Configuration, the DSL is parsed into a corresponding PipelineConfigurationObject.
These PipelineConfigurationObjects are then added together to create the Aggregated Pipeline Configuration.
flowchart LR
g[Global Configuration]
gc[Global Pipeline Configuration]
ac[Folder A Pipeline Config]
jc[Job Pipeline Config]
agg[Aggregated Pipeline Config]
g --> f_a[Folder A]
f_a --> j_b[Job B]
style j_b fill:#FFD580,stroke:#000000
gc <-.- g
ac <-.- f_a
jc <-.- j_b
gc -.->| + |ac
ac -.->| + |jc
jc -.->| = |aggNote
PipelineConfigurationObject overrides the plus operator so adding them together merges two configurations.
Pipeline Configuration Parsing¶
The Pipeline Configuration Syntax is a custom Domain-Specific Language (DSL).
The org.boozallen.plugins.jte.init.governance.dsl package has all the classes related to Pipeline Configuration parsing.
A Pipeline Configuration file is executed as a Groovy Script with a custom binding that makes the env variable available to the configuration file.
The Script is compiled using PipelineConfigurationBuilder as the script base class which overrides methodMissing, propertyMissing, and setProperty to build the PipelineConfigurationObject defined by the DSL.
This script is executed in the same groovy sandbox as Jenkins Pipelines to secure the script execution.