服务限流、降级、熔断-SpringCloud

本文所使用的组件：Nacos（服务中心和注册中心）、OpenFeign（服务调用）、Sentinel（限流、降级）、Hystrix（熔断）

项目结构：

1. service-provider：提供服务的微服务。

2. service-consumer：消费服务的微服务。

1. 添加依赖

在两个服务的pom.xml文件中添加Spring Cloud Alibaba、Nacos、Sentinel、Hystrix和OpenFeign的依赖。

<dependencies><!-- Spring Cloud Alibaba Nacos Discovery --><dependency><groupId>com.alibaba.cloud</groupId><artifactId>spring-cloud-starter-alibaba-nacos-discovery</artifactId></dependency><!-- Spring Cloud Alibaba Sentinel --><dependency><groupId>com.alibaba.cloud</groupId><artifactId>spring-cloud-starter-alibaba-sentinel</artifactId></dependency><!-- Hystrix --><dependency><groupId>org.springframework.cloud</groupId><artifactId>spring-cloud-starter-netflix-hystrix</artifactId></dependency><!-- Spring Cloud OpenFeign --><dependency><groupId>org.springframework.cloud</groupId><artifactId>spring-cloud-starter-openfeign</artifactId></dependency><!-- Spring Boot Web Starter --><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency>
</dependencies>

2. Docker-compose.yml文件

创建docker-compose.yml文件，启动Nacos和Sentinel。

version: '3'
services:nacos-server:image: nacos/nacos-server:latestcontainer_name: nacos-serverports:- "8848:8848"environment:- MODE=standalone- SPRING_DATASOURCE_PLATFORM=mysqlsentinel-server:image: apache/incubating-sentinel-dashboard:latestcontainer_name: sentinel-serverports:- "8080:8080"- "8719:8719"

• 部署在云服务中需要打开8080、8848端口，端口冲突可以更换端口。

3. service-provider（提供服务的微服务）

3.1 启动类

创建启动类并添加nacos注册客户端。

import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.cloud.openfeign.EnableFeignClients;@SpringBootApplication
@EnableFeignClients
public class ServiceProviderApplication {public static void main(String[] args) {SpringApplication.run(ServiceProviderApplication.class, args);}
}

3.2 控制器

创建控制器提供服务。

import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;@RestController
public class ServiceController {@GetMapping("/provider")public String provider() {return "Hello from Service Provider";}
}

4. service-consumer（消费服务的微服务）

4.1 OpenFeign客户端

创建Feign客户端用于调用service-provider。

import org.springframework.cloud.openfeign.FeignClient;
import org.springframework.web.bind.annotation.GetMapping;@FeignClient(name = "service-provider", fallback = ServiceFallback.class)
public interface ServiceClient {@GetMapping("/provider")String provider();
}@Component
public class ServiceFallback implements ServiceClient {@Overridepublic String provider() {// 服务降级逻辑return "Fallback response from Service Consumer";}
}

4.2 控制器

import com.alibaba.csp.sentinel.annotation.SentinelResource;
import com.alibaba.csp.sentinel.slots.block.BlockException;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RestController;@RestController
public class ConsumerController {private final ServiceClient serviceClient;public ConsumerController(ServiceClient serviceClient) {this.serviceClient = serviceClient;}@GetMapping("/consumer")@HystrixCommand(commandKey = "consumerCommand", fallbackMethod = "fallback")@SentinelResource(value = "serviceClientCall", blockHandler = "blockHandler")public String consumer() {return serviceClient.provider();}public String fallback() {// 熔断降级逻辑return "Service is down, hystrix fallback";}public String blockHandler(BlockException ex) {// Sentinel降级逻辑return "Service is blocked by Sentinel";}
}

• @HystrixCommand注解用于指定熔断器的命令键和降级方法，@SentinelResource注解用于指定资源名和降级方法。

5. application.yml

在两个服务添加application.yml文件。

spring:cloud:nacos:discovery:server-addr: nacos-server:8848config:server-addr: nacos-server:8848sentinel:transport:dashboard: sentinel-server:8080datasource:ds1:type: filefile:filePath: /path/to/your/sentinel/ruleshystrix:command:default:execution:isolation:strategy: THREADthread:timeoutInMilliseconds: 3000

• filePath:文件路径

6. 配置规则

在Sentinel Dashboard（http://localhost:8080）中添加限流规则：

资源名：serviceClientCall
限流模式：QPS
阈值：10

总结：上述为一套完整的服务治理流程，对于某些场景下可以使用guava框架去实现单体限流。主要学习思想，本文所使用的组件可替换为其他具有相关功能的组件。

不积跬步，无以至千里 --- xiaokai