How wjunicorn/coding653 works
Project Type and Purpose
• This is a multi-module full-stack, microservices-based platform repository.
It contains several domains:
• “Tag Server” – tagserver-ui (Vue.js frontend) and tagserver (Java Spring-Boot services: core, dao, web)
• “Metadata” – metadata-ui (Vue.js frontend) and metadata (Java Spring-Boot services: core, dao, web)
• “Query Middleware” – a set of Java modules (entrance-common, entrance-core, hiveserver2-entrance, presto-entrance, spark-livy-entrance, router-core, router-dao, router-web, router-cli, router-jdbc) implementing a gRPC/RPC-based distributed query router and engine clients
• “Cluster Manager” – manager-ui (Vue.js frontend) and platform-manager (Java Spring-Boot services: common, core, dao, web)
• “Auto-Deploy Installer” – platform-installer (Java console app + shell scripts for SSH orchestration)
• Dozens of chapter-based demo apps illustrating specific technologies
Key Architectural Patterns
• Clear frontend/backend separation: UIs built with Vue.js + D2Admin and webpack configs; RESTful API service layers behind them.
• Layered backend services:
• Web layer (Spring Boot controllers + SwaggerConfig)
• Core/service layer (business logic, scheduling with Drools, caching)
• DAO layer (MyBatis or JPA repositories)
• Microservices communicate via REST or gRPC within Query Middleware.
• External integrations:
• Databases (MySQL, ClickHouse)
• Caching (Redis)
• Service discovery/coordination (Zookeeper)
• Big-data systems (HDFS, YARN, Hive, Presto, Spark, Livy)