GitLab CI/CD Specialist case studies

GitLab CI/CD Specialist · Put your knowledge to work in context.

The GitLab CI/CD Specialist credential is aimed at engineers who build, operate, and troubleshoot delivery pipelines on GitLab rather than at general GitLab users. It targets DevOps engineers, platform and SRE staff, release engineers, and developers who own a project's .gitlab-ci.yml end to end - people expected to design pipeline structure, choose runners and executors, manage variables and secrets, and diagnose failures under time pressure. The GitLab CI/CD Specialist case studies are 5 cases and about 4,916 words.

Monorepo merge semantics: shared template, silent overrides, duplicate pipelines

Acme Pay runs a monorepo with 38 services. The root .gitlab-ci.yml includes templates/node.yml and then every services/<name /.gitlab-ci.yml through include: local - 39 include entries, none of them with rules .

908 words, about 5 min 2

Tag mismatch and executor drift on a 14-runner fleet

An e-commerce company runs GitLab self-managed with 14 group-level runners. All are Docker executor on EC2, each with concurrent = 4 in config.toml and limit = 6 in the runner settings.

982 words, about 5 min 3

Masked is not secret: two leaks and a four-month-old access key

payments-api deploys to AWS from a deploy:prod job using project CI/CD variables. AWS ACCESS KEY ID is Masked.

995 words, about 5 min 4

A 41-minute pipeline with a 3% cache hit rate and 1.9 GB of artifacts

web-frontend takes 41 minutes on 8-core autoscaling runners (six of them, on AWS). The graph is install ( npm ci , 9 min) → lint (3 min), unit (5 min), e2e (14 min, 340 Playwright tests run serially) → build (6 min, emits dist/ , 1.9 GB with source maps) → deploy-review .

1,000 words, about 5 min 5

A security gate that never blocks, a manual deploy from a feature branch, and exit 137

A regulated insurer runs GitLab Ultimate. The core-banking group attaches a compliance framework to every project, and the policy requires SAST, Secret Detection and Dependency Scanning before merge.

1,031 words, about 5 min