Skip to content

CLAUDE.md

Latest commit

 

History

History
111 lines (79 loc) · 3.52 KB

CLAUDE.md

File metadata and controls

111 lines (79 loc) · 3.52 KB
Raw

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

This is a DigitalOcean-specific adaptation of Google's Online Boutique microservices demo application. It consists of 12+ microservices written in Go, Python, Java, Node.js, and C# that communicate over gRPC, designed to showcase DOKS (DigitalOcean Kubernetes Service) and Database-as-a-Service integrations.

Repository Structure

  • src/ - Individual microservice implementations
  • helm/ - Helm chart for Kubernetes deployment
  • dev/ - Development configuration values
  • scratch/ - Temporary/scratch files

Services Architecture

Service Stack by Language:

  • Go: frontend, checkoutservice, productcatalogservice, shippingservice
  • Java: adservice (Gradle-based)
  • C#: cartservice (.NET 9.0)
  • Node.js: currencyservice, paymentservice
  • Python: emailservice, recommendationservice, loadgenerator, shoppingassistantservice

Key Service Dependencies:

  • adservice requires PostgreSQL database
  • cartservice requires Valkey/Redis database
  • frontend serves the web UI and orchestrates other services
  • loadgenerator simulates realistic user traffic

Deployment Configuration

The application supports two deployment modes via Helm:

  • Development mode (devDeployment: true): Uses in-cluster PostgreSQL and Valkey from Bitnami subcharts
  • Production mode (devDeployment: false): Requires external managed database connections via ConfigMaps and Secrets

Common Development Commands

Kubernetes/Helm Operations

# Deploy in development mode
helm install demo oci://ghcr.io/do-solutions/microservices-demo \
  --namespace boutique --create-namespace

# Deploy with custom values
helm install demo oci://ghcr.io/do-solutions/microservices-demo \
  --namespace boutique --create-namespace \
  -f dev/values.yaml

# Package Helm chart locally
helm package ./helm --version <version> --app-version <tag>

# Get frontend external IP
kubectl get svc frontend-external -n boutique -o jsonpath='{.status.loadBalancer.ingress[0].ip}'

Building Individual Services

# Java (adservice)
cd src/adservice && ./gradlew build

# C# (cartservice)
cd src/cartservice && dotnet build

# Go services (multiple)
cd src/<service> && go build

# Node.js services
cd src/<service> && npm install

# Python services
cd src/<service> && pip install -r requirements.txt

Running Tests

# Java tests
cd src/adservice && ./gradlew test

# C# tests
cd src/cartservice && dotnet test

# Go tests (where available)
cd src/<service> && go test ./...

CI/CD System

The repository uses GitHub Actions with a unified build system:

  • All services are built and tagged with the same version (v1.YYYYMMDD.HHMMSSfff)
  • Docker images are published to GitHub Container Registry (GHCR)
  • Helm chart is packaged and pushed as an OCI artifact
  • Workflow triggers on changes to src/** or helm/** on the main branch

External Database Configuration

For production deployments, create these Kubernetes resources before deployment:

AdService PostgreSQL:

  • ConfigMap: adservice-database-configuration with DB connection details
  • Secret: adservice-database with postgres-password key

CartService Valkey/Redis:

  • Secret: cartservice-database with connectionString key

Container Images

All service images are available at: ghcr.io/do-solutions/microservices-demo-<service>:<tag>

Services are containerized with optimized Dockerfiles in each service directory.