ProfessionalCI/CD modernisation
Migrating Terraform pipeline templates from Azure DevOps to GitHub Actions
Translated Azure DevOps pipeline templates for Terraform-based Infrastructure as Code into GitHub Actions workflows, preserving equivalent build and deployment logic.
- Role
- Cloud DevSecOps Engineer — pipeline migration
- Context
- ALIANDO · Cloud Platform area
- Period
- 2026
01Overview
Pipeline templates encode a team’s delivery standards. Moving them from Azure DevOps to GitHub Actions means re-expressing those standards in a different execution model without changing what gets built or deployed. I carried out that translation for Terraform-based IaC workflows.
02Context
Infrastructure in the Cloud Platform area is delivered through Terraform executed by CI/CD pipelines. Pipeline templates already existed in Azure DevOps; the requirement was to provide the same capability in GitHub Actions.
03Problem statement
Azure DevOps and GitHub Actions share concepts but not semantics — templates, stages, variable scoping and approvals are modelled differently. A literal port risks subtle behavioural changes in infrastructure deployments, where a different variable value or a skipped step can change real resources.
04My responsibilities
- Analysed the existing Azure DevOps pipeline templates for Terraform IaC.
- Translated them into GitHub Actions workflows.
- Preserved equivalent build and deployment logic across the migration.
05Architecture overview
The migration maps each Azure DevOps construct to its GitHub Actions counterpart. The diagram shows that conceptual mapping between the two platforms’ building blocks; the Terraform logic executed inside the pipelines stays the same.
- Delivery & automation
- Code & configuration
Conceptual view of the platform constructs involved; it does not reproduce the client’s pipelines.
Text description of the diagram
Azure DevOps pipeline templates map to GitHub Actions reusable workflows; stages, jobs and steps map to jobs and steps; pipeline variables map to workflow inputs, variables and secrets; environments and approvals map to GitHub environments. Both run the same Terraform logic.
Flows
- Pipeline templates → Reusable workflows
- Stages · jobs · steps → Jobs · steps
- Variables · variable groups → Inputs · variables · secrets
- Environments · approvals → Environments · protection rules
Components
Pipeline templates · Stages · jobs · steps · Variables · variable groups · Environments · approvals · Reusable workflows · Jobs · steps · Inputs · variables · secrets · Environments · protection rules · Same Terraform logic
06Technology stack
- Source platform
- Azure DevOps Pipelines
- YAML templates
- Target platform
- GitHub Actions
- Reusable workflows
- Workload
- Terraform
- Infrastructure as Code
07Implementation approach
- Inventory the existing templates and the behaviour each one guarantees — inputs, ordering, conditions and deployment targets.
- Re-express each template in GitHub Actions terms, keeping the Terraform commands and their order unchanged.
- Treat equivalence as the acceptance criterion — the migrated workflows must build and deploy the same way as the originals.
08Challenges & engineering decisions
Challenge
Different execution models — Azure DevOps stages and templates do not map one-to-one onto GitHub Actions workflows and jobs.
Decision
Map by intent rather than by syntax — identify what each template guarantees and reproduce that guarantee with the closest GitHub Actions construct.
Challenge
In infrastructure pipelines, a behavioural difference is a change to real resources.
Decision
Keep the Terraform workflow itself untouched and limit the migration to the orchestration layer around it.
09Security considerations
- Secrets and variables must keep their scoping when moved between platforms — a variable that was environment-scoped must not become globally readable.
- Deployment controls on sensitive environments need an explicit equivalent on the target platform.
10Operational considerations
- Equivalent logic means operators keep the same mental model of how infrastructure is delivered.
- Reusable workflows keep delivery standards centralised instead of copied into every repository.
11Results & impact
- Azure DevOps pipeline templates for Terraform IaC translated to GitHub Actions.
- Equivalent build and deployment logic preserved.
12Lessons learned
- A pipeline migration is a behaviour-preservation exercise; syntax is the easy part.
- Separating orchestration from the infrastructure logic it runs makes CI/CD platforms interchangeable.