goodsKill

Why we included this project

goodsKill is a teaching codebase for high-concurrency Java backends, and it earns that label by wiring a real flash-sale flow through a surprisingly broad stack: Dubbo RPC, Nacos, Spring Cloud Gateway, Sharding-JDBC for sharding, Seata for distributed transactions, and a state machine for managing state changes. Instead of reading separate blog posts about each piece, you can trace how they fit together in one codebase. It also ships ten concurrency strategies, from synchronous locks and Redis or ZooKeeper distributed locks to Kafka and RabbitMQ queues and atomic updates, so the trade-offs sit side by side. The AI part stays deliberately small: a Spring AI chatbot that can trigger simulated purchases, a concrete example of slotting an AI agent into an existing Java microservice without rebuilding around it. The OpenTelemetry setup bundled with Grafana's LGTM stack is a practical bonus for anyone studying tracing and logging. Just know this is a learning reference, not something you would deploy as-is, but for engineers studying distributed systems or preparing for high-throughput backend interviews it covers a lot of ground.