Goals:

1. Build a containerized application using the microservices architecture
2. Deploy the application on the cloud with Kubernetes
3. Configure the required infrastructure with Terraform's IaC approach

Non-goals:

1. Create a useful service or API

Structure

Name	Framework	Information
api-gateway	Spring WebFlux, Spring Cloud Gateway	Centralizes communication between services and exposes them
user-service	Spring Web	Responsible for information about some "users"
content-service	Spring Web	Responsible for information about some "content"
recommendation-service	Spring WebFlux	Aggregates user and content data and produces content recommendations for users

Note: Each service is intentionally a separate Gradle project, rather than a module of a parent project. That is because microservices should ideally be decoupled. However, for convenience, they are all put in this single repository.

Deploying

Create a Google Cloud project

Create a new project in the GCP Console. Enable the Compute Engine API, Kubernetes Engine API, Cloud Build API and Artifact Registry API.

Setup the Google Cloud client

gcloud init
gcloud auth application-default login
gcloud components install kubectl
gcloud components install gke-gcloud-auth-plugin

Prepare the infrastructure with Terraform

Configure the variables in terraform.tfvars, then run

terraform init
terraform apply

Build Docker images

For each service run

gcloud builds submit --tag <region>-docker.pkg.dev/<project>/<repo>/<service_name> .

Create Kubernetes deployments and services

For each service, configure the image location in the deployment.yaml file, then run

kubectl apply -f deployment.yaml
kubectl apply -f service.yaml

Sources

• Deploy an app in a container image to a GKE cluster
• Provision a GKE Cluster
• Configure private services access
• Deployments
• Service discovery and DNS
• Inter-cluster communication between microservices on Kubernetes
• kubectl authentication in GKE