Pipeline Syntax (.gitlab-ci.yml)

Overview

Write clear .gitlab-ci.yml using variables, workflow and job rules, and understand needs vs dependencies for ordering and artefact flow.

GitLab reads .gitlab-ci.yml (or an alternate CI config path) to define pipelines. Prefer rules over legacy only / except. Use workflow:rules to decide whether a pipeline is created at all. Use needs for DAG edges; use dependencies to control which job artefacts download.

This is a core tutorial in Module 4 · Pipeline Syntax of the REBASH Academy GitLab CI/CD for Cloud & DevOps Engineers series — written for Cloud, DevOps, Platform, and SRE engineers.

Prerequisites

• GitLab Runners and Executors

Learning Objectives

By the end of this tutorial, you will be able to:

• Structure stages, default, variables, and jobs
• Prefer rules over only/except
• Control pipeline creation with workflow
• Contrast needs (ordering) with dependencies (artefacts)

Architecture

This topic’s control points and relationships are shown below.

Theory

What it is

Top-level keys commonly include stages, default, variables, workflow, include, and job names. A job needs at least a script (or trigger / trigger:include patterns covered later). Predefined variables such as $CI_COMMIT_SHA, $CI_PIPELINE_SOURCE, and $CI_DEFAULT_BRANCH are injected by GitLab — prefer them over hard-coded branch names where possible.

Keyword	Role
variables	Pipeline or job env (plus UI/group vars)
rules	When a job is added to the pipeline
workflow:rules	Whether any pipeline is created
needs	Start job early; optional artefact download
dependencies	Which prior jobs’ artefacts to fetch (stage model)

Why it matters

Ambiguous YAML creates ghost pipelines (double branch + MR pipelines), skipped deploy jobs, or jobs that download every artefact and waste minutes. Precise rules and lean dependencies cut cost and confusion. Reviewers can reason about production risk from the YAML alone.

How it works

1. GitLab loads YAML (and includes).
2. workflow:rules decide if this event creates a pipeline.
3. For each job, rules decide inclusion; variables expand.
4. Default graph: all jobs in a stage finish before the next stage starts.
5. needs can start a job as soon as listed jobs finish (DAG), optionally with artifacts: true/false.
6. Without dependencies, jobs may download all artefacts from previous stages; set dependencies: [] or an explicit list to stay lean.

Migrate away from only / except — they still work but compose poorly with complex sources.

Key concepts and comparisons

Approach	Prefer when
Stage ordering only	Simple linear build → test → deploy
needs DAG	Long pipelines; parallel fans; skip empty stages
dependencies	Control artefact download under stage ordering

Legacy	Modern
only: [main]	rules: [{ if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH }]
except: [schedules]	rules with when: never for that source

Common pitfalls

• Combining only and rules on the same job — pick one model (rules).
• Forgetting workflow and paying for duplicate MR + branch pipelines.
• Using needs but still assuming stage barriers apply the same way.
• Putting secrets in YAML variables: — use masked/protected UI vars or OIDC (Module 6).

Hands-on Lab

Create a workspace for this tutorial.

mkdir -p ~/rebash-gitlab/module-04 && cd ~/rebash-gitlab/module-04

Focus: exercise rules, needs, and parallel matrix syntax

Step 1 – Write advanced syntax pipeline

cat > .gitlab-ci.yml << 'EOF'
workflow:
  rules:
    - if: $CI_PIPELINE_SOURCE == "merge_request_event"
    - if: $CI_COMMIT_BRANCH
stages: [lint, test, build]
lint:
  stage: lint
  image: alpine:3.20
  script: ["echo linting $CI_COMMIT_SHA"]
unit:
  stage: test
  image: python:3.12-alpine
  parallel:
    matrix:
      - PYTHON: ["3.11", "3.12"]
  script: ["echo matrix python=$PYTHON"]
build:
  stage: build
  image: alpine:3.20
  needs: [lint, unit]
  rules:
    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH
    - if: $CI_PIPELINE_SOURCE == "merge_request_event"
      when: manual
  script: ["mkdir -p dist && echo ok > dist/marker.txt"]
  artifacts:
    paths: [dist/]
EOF

Step 2 – Inspect keywords

grep -E '^(workflow:|parallel:|needs:|rules:)' -n .gitlab-ci.yml
grep -A3 'needs:' .gitlab-ci.yml

Final step – Cleanup note

# Keep ~/rebash-gitlab/ for later tutorials

Validation

• Lab commands run under ~/rebash-gitlab/module-04/
• You can explain each Theory section in your own words
• You used modern tooling where it applies to this topic
• You can describe one production failure mode for this topic

Code Walkthrough

Production practice for Pipeline Syntax (.gitlab-ci.yml) always combines:

1. Inspect before you change (status, plan, logs, dry-run)
2. Prefer reversible, documented changes (Git, IaC, drop-ins, version pins)
3. Capture evidence (command output, pipeline logs) for handovers
4. Prefer current tools and APIs over legacy shortcuts
5. Least privilege — escalate credentials only when required

Keep runbooks short enough to follow under pressure. Automate checks; keep humans for judgement.

Security Considerations

• Treat credentials and tokens for gitlab as privileged — never commit them
• Prefer short-lived auth (OIDC, roles, SSO) over long-lived keys
• Validate blast radius before apply/deploy/delete operations
• Restrict who can approve production changes
• Collect audit logs; limit who can read sensitive traces

Common Mistakes

Combining only and rules on the same job — pick one model (rules).

Validate assumptions against the Theory section and official docs before changing production.

Forgetting workflow and paying for duplicate MR + branch pipelines.

Lab shortcuts (open security groups, admin roles, skip approvals) must not ship unchanged.

Changing production without a rollback path

Always know how to revert (previous artefact, prior release, state rollback, DNS failback).

Best Practices

• Encode Pipeline Syntax (.gitlab-ci.yml) changes as code and review them in pull requests
• Pin versions (images, modules, actions, provider plugins)
• Separate environments with clear promotion gates
• Alert on symptoms with runbooks attached
• Destroy lab resources; tag everything with owner and expiry where possible

Troubleshooting

Symptom	Likely cause	Fix
Auth / permission denied	Wrong identity, policy, or scope	Check caller identity, roles, and least-privilege policies
Timeout / no route	Network, DNS, security group, or endpoint	Trace path, DNS, and allow-lists before retrying
Drift / unexpected plan	Manual change or wrong state/workspace	Reconcile desired vs actual; avoid click-ops on managed resources
Pipeline/job red	Flaky step, cache, or missing secret	Read failing step logs; bisect recent workflow/config changes
Cost spike	Idle load balancer, NAT, oversized compute	Inventory billable resources; stop/delete labs promptly

Summary

Pipeline Syntax (.gitlab-ci.yml) is essential for Cloud and DevOps engineers working with gitlab. Practise the lab until the inspection and change path is muscle memory, then continue the track.

Interview Questions

1. What does needs change compared with stage-only ordering?
2. How do rules differ from only/except, and why prefer rules?
3. When is parallel matrix the right tool?
4. How can a malicious merge request abuse overly broad rules?
5. What is workflow rules for at the pipeline level?

Sample answer — question 2

Confirm whether the job was skipped by rules, blocked by needs on a failed dependency, or never created because workflow rules prevented the pipeline.

Sample answer — question 4

Prefer least-privilege rules: block pipelines from untrusted forks where secrets are available, keep production deploys manual on the default branch.

Related Tutorials

• Course overview
• Pipeline Design: DAGs and Includes

References

• CI/CD YAML syntax
• Job rules