Platform Delivery Blueprint (Terraform + AWS + Kubernetes)

Project scope

I built an end-to-end cloud delivery blueprint combining Terraform for AWS, a containerized Node.js service, Kubernetes manifests and GitHub Actions. It brings infrastructure provisioning, deployment automation and runtime health checks into one personal engineering project.

View source on GitHub View GitHub Actions workflow

Architecture

The Terraform-managed AWS foundation includes a VPC, public/private subnets, NAT, IAM, EKS control plane, managed node group, ECR repository, CloudWatch log group, and ACM certificate flow.
The application layer is a Dockerized Node.js API with /healthz, /readyz, and Prometheus-style /metrics endpoints, plus structured stdout logs.
Kubernetes deployment patterns include rolling updates, readiness and liveness probes, NLB service exposure, HPA thresholds, and a PodDisruptionBudget.
CI/CD validation and deployment are handled in GitHub Actions with remote Terraform state, OIDC-based AWS access, and Route53 DNS automation.

Architecture view

GitHub push / pull request

↓

GitHub Actions

Node.js tests
Kubernetes manifest render
Terraform fmt + validate
OIDC deploy job

↓

AWS VPC + subnets + NAT + IAM + EKS + ECR + CloudWatch + ACM + SSM

↓

Kubernetes deployment + load balancer service + HPA + PDB

↓

Node.js API with health, readiness, metrics, structured logs, DNS, and HTTPS

Repository implementation

/platform/aws-platform-showcase/bootstrap provisions remote state storage and the GitHub OIDC deployment role.
/platform/aws-platform-showcase/terraform contains the infrastructure code.
/platform/aws-platform-showcase/k8s contains the deployable Kubernetes manifests.
/platform/aws-platform-showcase/app contains the sample service and tests.
/.github/workflows/platform-showcase.yml validates the showcase on relevant changes.

Runtime configuration

The deployment manifest defines readiness, liveness and startup probes.
Rolling update settings, replica limits and HPA thresholds are defined in the manifests.
Metrics are exposed in Prometheus format and service/pod annotations support scrape-based monitoring.
Terraform defines an SSM parameter path for configuration; the sample application reads its settings from environment variables.
ECR image scanning, ACM certificate issuance, CloudWatch log retention, and OIDC-based deployment access are defined in the infrastructure layer.

Deliverables & deployment status

Infrastructure definitions, deployment manifests and a sample API with tests are available in this repository.
The deployment workflow requires an AWS account, environment configuration and an explicit deployment run.
Live deployment status, uptime and load-test measurements are not reported for this personal project.