How revandpratama/reflect-be works
Project type and purpose
• This repo is a Go-based microservices system.
• It consists of three independently deployable services—auth-service, logging-service, and main-app—plus a shared “shared/proto” folder for gRPC contracts and a “deployments” folder for Docker Compose orchestration.
• Purpose: main-app is the primary API gateway and business logic layer (exposes REST/gRPC), auth-service handles user authentication/authorization, logging-service processes log events (via Kafka).
File structure
• Top-level directories:
– auth-service: hexagonal (adapter, config, helper, internal/{controller,services,repositories,entities,dto}, generatedProtobuf)
– logging-service: adapter (Kafka), server entrypoint
– main-app: rich layered structure (adapter for gRPC, Kafka, MinIO, PostgreSQL, Redis, REST; cmd; config; errorhandler; helper; internal/{handlers,middleware,repositories,services,routes,dto,entities,generatedProtobuf}; types)
– shared/proto: .proto definitions (auth.proto) shared by services
– deployments: docker-compose setups for local and CI/CD
• Architectural clues:
– Hexagonal/Clean layers in each service
– Event-driven integration via Kafka
– Sync RPC via gRPC (auth proto)
– Polyglot persistence: PostgreSQL, Redis, MinIO object store
README (empty here)
• No explicit architecture section—derive from file layout and naming conventions.