liveBook · Manning

Dynamic release pipeline

As explained in Effective Platform Engineering

A dynamic release pipeline is an automation pipeline that generates its workflow at runtime from the currently defined deployment targets and roles.

A dynamic release pipeline generates deploy and test workflow at runtime from a global definition of current roles and deployment targets. It replaces hard-coded cluster lists, reducing brittleness when one orchestration pipeline serves more than 20–30 clusters. section 9.1.1

A practical implementation follows this pattern:

  • Maintain globally available data listing roles and clusters per role. section 9.1.1
  • Generate deploy and test stages before execution using the current definition. section 9.1.1
  • Use the pipeline tool’s matrix capabilities where they fit, without making implementation details the design. section 9.1.1

Regional role expansion

If a role contains four clusters across four regions, automation creates matching deploy and test stages. A smaller configuration can experiment with development and quality-assurance roles, each containing regional cluster instances. section 9.1.1

test:
  filter: "*on-push-main"    #1
  deploy:    #2
    - platform-dev
    - platform-qa
  roles:
    platform-dev:    #3
      deploy:    #4
        - platform-dev-i01-aws-us-west-2
        - platform-dev-i01-aws-eu-west-1
      instances:
        platform-dev-i01-aws-us-west-2:    #5
          aws_region: us-west-2
          aws_account_id: '10100000000'
        platform-dev-i01-aws-eu-west-1:
          aws_region: eu-west-1
          aws_account_id: '10100000000'
    platform-qa:
      deploy:
        - platform-qa-i01-aws-us-west-2
        - platform-qa-i01-aws-eu-west-1
      instances:
        platform-qa-i01-aws-us-west-2:
          aws_region: us-west-2
          aws_account_id: '10100000000'
        platform-qa-i01-aws-eu-west-1:
          aws_region: eu-west-1
          aws_account_id: '10100000000'

The configuration listing separates pipeline triggers, roles, deployment targets, and per-instance regional account data—the inputs used to generate runtime workflow. section 9.1.1

How Effective Platform Engineering covers it

How deep it goes
Defines runtime generation and global configuration, but leaves the complete implementation to the pipeline tool and an experiment.
Examples
Uses multi-region cluster roles, regional instance metadata, and an exercise generating deploy and test stages.
What you can do after
Build it from scratch by defining role data and generating matching deploy and test stages at runtime.
How it connects
Connects matrix builds, pipeline automation, deployment roles, clusters, and global definitions.
Cover of Effective Platform Engineering
liveBookmanning.com
Read Effective Platform Engineering on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Effective Platform Engineering. The section links point to the parts of the book it draws on.