Skip to content

Deployment Pipeline

Home › Explanation › Continuous Delivery › Core Concepts › Pipeline

Introduction

The Deployment Pipeline is the automated, repeatable process that takes code from commit to production (or artifact registry).

Every change flows through the pipeline, which validates quality, builds artifacts, and delivers software to its destination.

Deployment Pipeline Legend

Definition

Deployment Pipeline: The automated process that validates and delivers changes through the CD Model stages.

Key Characteristics:

  • Triggered by changes to trunk
  • Implements the 12 stages of the CD Model
  • Provides automated quality gates
  • Generates immutable artifacts
  • Collects evidence and metrics

One Pipeline Per Deployable Module

Each deployable module has its own pipeline instance:

  • Polyrepo: One pipeline per repository
  • Monorepo: Multiple pipelines in one repository (one per deployable module)

Pipelines are triggered when changes affect their deployable module. In a monorepo, change detection determines which pipelines run.

Trunk Commit
    │
    ├─► Module A changed? ─► Run Pipeline A
    ├─► Module B changed? ─► Run Pipeline B
    └─► Module C changed? ─► Run Pipeline C

Pipeline Triggers

Pipelines can be triggered by:

Trigger When Purpose
Commit Code merged to trunk Main validation and artifact creation
Pull Request PR opened/updated Pre-merge validation
Schedule Cron (daily, weekly) Dependency updates, security scans
Manual User-initiated Releases, hotfixes
Tag Git tag created Release candidate creation

The 12 Stages

The pipeline implements the CD Model's 12 stages:

Development Stages (1-7):

  1. Authoring - Developer creates changes on topic branch
  2. Pre-commit - Local validation before push
  3. Merge Request - PR validation and review
  4. Commit - Build artifacts on trunk merge
  5. Acceptance Testing - Validate in PLTE
  6. Extended Testing - Performance, security, compliance
  7. Exploration - Stakeholder validation in Demo

Release Stages (8-12):

  1. Start Release - Create release candidate
  2. Release Approval - Manual or automated gate
  3. Production Deployment - Deploy to live
  4. Live - Monitor in production
  5. Release Toggling - Feature flag control

See CD Model Stages for detailed stage descriptions.

Evidence Collection

The pipeline collects evidence at each stage for traceability and compliance:

Stage Evidence Collected
Pre-commit Lint results, unit test results
Commit Build logs, artifact checksums, SBOM
Acceptance Testing Test reports, coverage metrics
Extended Testing Performance benchmarks, security scan results
Production Deployment Deployment logs, health check results
Live Monitoring dashboards, incident records

Evidence enables:

  • Audit trails for compliance
  • Debugging production issues
  • Release quality metrics
  • Continuous improvement

Pipeline Variants

The pipeline structure varies based on the CD Model variant:

Release Assurance (RA):

  • Manual approval gate at Stage 9
  • Release branches for stabilization
  • Explicit release candidate management

Continuous Deployment (CDe):

  • Automated approval based on quality gates
  • Direct trunk-to-production flow
  • Feature flags for release control

See CD Variants for detailed comparison.

Pipeline Implementation

Pipelines are typically implemented using:

  • GitHub Actions - .github/workflows/
  • Azure DevOps - azure-pipelines.yml
  • GitLab CI - .gitlab-ci.yml
  • Jenkins - Jenkinsfile

The pipeline definition is stored in the trunk alongside the code it builds, following "Everything as Code" principles.

Next Steps

References

Tutorials | How-to Guides | Explanation | Reference

You are here: Explanation — understanding-oriented discussion that clarifies concepts.