Video: https://youtu.be/QqitAgA-RXw

Backend Source Code: https://github.com/zli78122/gulimall
Frontend Source Code: https://github.com/zli78122/gulimall_renren-fast-vue

• Developed a distributed e-commerce system based on Spring Boot and Spring Cloud, used Docker to manage multiple middleware and Nginx for dynamic and static separation of resources, reverse proxy, and load balancing.
• Introduced a complete set of microservice governance solutions: Nacos as a registration and configuration center, Gateway as a gateway, Feign for remote call, Ribbon for load balancing, Sentinel for flow protection, Sleuth and Zipkin as a tracing system.
• Solved most of the problems faced by a highly concurrent distributed system: Spring Cache and Redis as distributed cache, Elasticsearch for faster product retrieval time, Spring Session for session data sharing, thread pool and asynchronous task for stability and performance.
• Implemented generating/cancelling order and locking/unlocking inventory using RabbitMQ delayed queue based on BASE theory and Flexible Transaction - message reliability and eventual consistency of transactions in a distributed system.
• Completed product flash sale using Redisson distributed lock – Semaphore and MQ, which can process 50,000 QPS in one Tomcat server.

Understanding the following four concepts is crucial for you to understand the relationship between these entities.

• Sales Attribute: The different attribute values of the sales attributes of a product will lead to differences in the sales price of this product. e.g. iPhone 12 Pro Max 256GB and iPhone 12 Pro Max 512GB have a different sales price. So "capacity" is a sales attribute
• Regular Attribute: It will not affect the price of the product, like the length, weight, ppi of iPhone 12 Pro Max.
• SPU (Standard Product Unit): It is a type of goods, such as a type of iPhone, like iPhone 12 Pro Max.
• SKU (Stock Keeping Unit): It is a specific product unit based on the SPU, like iPhone 12 Pro Max Gold+256GB, iPhone 12 Pro Max Gold+512GB, iPhone 12 Pro Max Blue+256GB, iPhone 12 Pro Max Sliver+512GB. i.e. SKU is a combination of various Sales Attributes (color + capacity) of the SPU.

(Delayed Queue, BASE Theory, Distributed Flexible Transaction)

• Locking Inventory: After the order is generated and before the user pays, the product purchased by current user will be locked in the inventory, which means that during this period, the product only belongs to the current user, and any other users can not purchase this product.
• Unlocking Inventory: When the order is paid or cancelled, the system will unlock the inventory.

(Distributed Lock - Semaphore, MQ)

• Flash Sale: Online store offers substantial discounts or promotions for a very short period of time. The main goals of a flash sale strategy are to get online shoppers to impulse buy, to increase short-term sales, or to sell your surplus stock. So a flash sale means a very high amount of concurrency, which may reach millions of concurrency.

IntelliJ IDEA, Visual Studio Code, Kibana

MySQL, Redis, Google Cloud Storage

Spring, Spring MVC, MyBatis/MyBatis-Plus, Spring Boot, Spring Cache, Spring Session
Spring Cloud : Nacos, Ribbon, Feign, Gateway, Seata, Sentinel, Sleuth, Zipkin
Thymeleaf, Elasticsearch, RabbitMQ
Redisson, CompletableFuture, Alibaba Cloud SMS, OAuth 2.0, Alipay, Cron

HTML, CSS, JavaScript, jQuery, XML, JSON
ES6, Axios, element-ui, Node.js, Webpack, Vue.js

Linux, Vagrant, Docker, Nginx

Maven, Git/GitHub, Apache JMeter

vagrant init centos/7  
vagrant up  
vagrant ssh  
exit;  
vagrant reload  

MySQL
Redis
Elasticsearch
Kibana
Nginx
RabbitMQ

Spring Cloud Alibaba - Nacos         registration center
Spring Cloud Alibaba - Nacos         configuration center
Spring Cloud         - Ribbon        load balancing
Spring Cloud         - Feign         remote call
Spring Cloud Alibaba - Sentinel      flow protection
Spring Cloud         - Gateway       gateway
Spring Cloud         - Sleuth        tracing system
Spring Cloud Alibaba - Seata         distributed transaction

Spring Cache + Redis, solve these three problems
    (1). Cache Penetration
    (2). Cache Crash
    (3). Cache Breakdown
Local Lock: synchronized(this);
Distributed Lock - Redisson
    (1). Reentrant Lock
    (2). ReadWrite Lock
    (3). Semaphore
    (4). CountDownLatch
Spring Cache:
    @Cacheable   : triggers cache population
    @CacheEvict  : triggers cache eviction
    @CachePut    : updates the cache without interfering with the method execution
    @Caching     : regroups multiple cache operations to be applied on a method
    @CacheConfig : shares some common cache-related settings at class-level

Callable + FutureTask & Thread Pool
CompletableFuture ≈ Promise in Frontend
Customized Thread Pool

Data Structure: Map<String key1, Map<String key2, CartItem>>
    key1: Shopping Cart Identifier
        For a Online Shopping Cart， key1 = "gulimall:cart:[userId]"   e.g. key1 = "gulimall:cart:1"
        For a Offline Shopping Cart，key1 = "gulimall:cart:[UUID]"     e.g. key1 = "gulimall:cart:6cd5b6f7-4349-48b9-bb7d-feccadbecae1"
        Note: Online Shopping Cart means login user, Offline Shopping Cart means unidentified user
    key2 : Product Identifier

Guarantee idempotence of interface using token.
Lua Script
Distributed Transaction (CAP & BASE)
    Solution 1. Spring Cloud Alibaba Seata
    Solution 2. RabbitMQ Delay Queue
Online Payment: Alipay

Service Single Responsibility: Dedicated and Independent Microservices
Random Code for Product Flash Sale - Token
Distributed Lock - Semaphore
RabbitMQ