Cloud Native DevOps Pipeline
Built an end-to-end DevOps pipeline from source code to Kubernetes deployment, covering CI/CD automation, containerization, infrastructure provisioning, and monitoring in a cloud-native workflow.
Overview
This project was built to understand and implement the full delivery flow of a modern DevOps environment: from writing code and building Docker images to provisioning infrastructure, deploying on Kubernetes, and monitoring the running application.
Architecture
This project follows an end-to-end DevOps workflow: source code is pushed to GitHub, automated through GitHub Actions, container images are stored in GHCR, infrastructure is provisioned with Terraform on Hetzner, and the application is deployed to a k3s Kubernetes cluster with monitoring through Prometheus and Grafana.
What I Built
- Containerized the application with Docker and prepared it for repeatable deployment.
- Built a GitHub Actions workflow for build, validation, image push, and deployment steps.
- Provisioned infrastructure using Terraform on Hetzner Cloud.
- Configured Kubernetes resources including Deployment, Service, Ingress, ConfigMap, and Secret.
- Applied liveness and readiness probes to improve deployment reliability.
- Configured Prometheus and Grafana for monitoring and dashboard visibility.
Challenges & Solutions
One of the main challenges was connecting several DevOps components into one coherent workflow rather than treating them as isolated tools. The project required decisions about image flow, infrastructure ownership, deployment order, and service exposure in Kubernetes.
-
Challenge: Making deployment repeatable across infrastructure and application layers.
Solution: Separated infrastructure provisioning with Terraform from application deployment through Kubernetes manifests and CI/CD steps. -
Challenge: Avoiding unstable rollouts and hidden runtime failures.
Solution: Added readiness and liveness probes so deployments could be validated by Kubernetes. -
Challenge: Observing the system after deployment, not just completing deployment.
Solution: Integrated Prometheus and Grafana to monitor health and metrics after rollout.
Result
The project gave me practical experience in building a complete DevOps delivery path rather than only using individual tools. It improved my understanding of how CI/CD, infrastructure as code, Kubernetes deployment, and monitoring work together in one system.