DevOps and CI/CD

I get your website running on your own VPS or in the cloud — to the modern standards you already expect from development: everything in Docker, infrastructure described as code, and every change deployed through a CI/CD pipeline that tests it first.

Deploying to a VPS or the cloud

Based on traffic and budget I pick what makes sense: a virtual server or AWS. I build the environment in Docker, so it is reproducible and identical in development and production. I prepare it alongside your current setup and switch over once it is verified.

CI/CD and tests in the pipeline

Uploading changes to the server by hand is the most common source of outages. I set up a CI/CD pipeline (GitLab CI, GitHub Actions) that runs on every change:

  • Code checks – coding standards and static analysis (PHPStan, ESLint)
  • Automated tests – unit and integration tests (PHPUnit, Jest) and end-to-end tests in a browser (Playwright)
  • Dependency checks – known vulnerabilities in the libraries you use
  • Deployment – the same procedure every time, with a way back if something goes wrong

Infrastructure built to fit

  • Containerization: Docker images, Docker Compose, and container optimization
  • Cloud infrastructure: AWS (EC2, S3, RDS, Lambda)
  • Infrastructure as Code: Terraform and CloudFormation for reproducible infrastructure
  • Development environments: DDEV, so every developer works on the same configuration
  • Secure operation: secrets and access kept out of the repository, blue-green deployments so a change can be rolled back, and a recovery plan that has actually been tried

Ongoing care after launch – security patches, backups and their restore – is described on the Website Maintenance page.

Technologies & tools

  • Containers: Docker, Docker Compose
  • CI/CD: GitLab CI, GitHub Actions
  • Testing: PHPUnit, Jest, Playwright, PHPStan
  • IaC: Terraform, CloudFormation
  • Cloud: AWS
  • VPS: Linux, Docker

What you get

  • A site on infrastructure you control
  • Deployments with no manual steps, tested before they reach production
  • An environment that can be rebuilt from source at any time

Frequently Asked Questions

Can the site be moved without a long outage?
Yes. I prepare the environment in parallel, verify everything works, and only then switch DNS.

Do we need our own server, or is cloud enough?
It depends on traffic and budget. Smaller sites are usually fine on a VPS; larger or spiky workloads justify AWS. I will recommend one once I know your numbers.

What if we end the engagement?
The infrastructure is described as code and runs on your accounts. I hand over documentation and access — nothing is locked in with me.

I run your website on your infrastructure, with every change tested before it goes live. Contact me on LinkedIn or use the form below.

Ask for Service →