Spring Cloud Ribbon

• 客户端负载均衡组件Ribbon 负载均衡就是当client端发送请求得时候会均衡得分发到每个server端

创建spring-cloud-02-ribbon-eureka 就是个服务端 参考前面得copy过来的

创建spring-cloud-02-ribbon-client1 项目

• 引入pom.xml

<dependency>
            <groupId>org.springframework.cloud</groupId>
            <artifactId>spring-cloud-starter-eureka</artifactId>
</dependency>

• UserController

@RestController
public class UserController {


    @RequestMapping(value = "/getUser", method = {RequestMethod.GET})
    public User getUser(@RequestParam("id") String id) {
        System.err.println("client1 GET-------------" + id);
        return new User(id, "20", "张三");
    }
}

• application.properties

spring.application.name=client-service
server.port=1003
server.context-path=/
eureka.client.service-url.defaultZone=http://eureka1:8003/eureka/
spring.output.ansi.enabled=always

• copy client1项目一份

创建 spring-cloud-02-ribbon-request

• 引入pom.xml

<dependency>
        <groupId>org.springframework.cloud</groupId>
        <artifactId>spring-cloud-starter-eureka</artifactId>
</dependency>
<dependency>
        <groupId>org.springframework.cloud</groupId>
        <artifactId>spring-cloud-starter-ribbon</artifactId>
</dependency>

• application.properties

spring.application.name=request-service
server.context-path=/
server.port=1005
eureka.client.service-url.defaultZone=http://eureka1:8003/eureka/
spring.output.ansi.enabled=always

• 启动所有 request访问http://localhost:1005/get 当请求得次数一样时，会发现分别在client1和client2端会打印出响应得次数，每个都打印出一半
• 这就是负载均衡 Ribbon

在这里在介绍下retry重试机制

spring-cloud-02-ribbon-retry

• pom.xml

<dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-eureka</artifactId>
</dependency>
<dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-ribbon</artifactId>
</dependency>
<dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-hystrix</artifactId>
</dependency>
<dependency>
    <groupId>org.springframework.retry</groupId>
    <artifactId>spring-retry</artifactId>
</dependency>

• application.properties配置

spring.application.name=retry-service
server.context-path=/
server.port=1006
eureka.client.service-url.defaultZone=http://eureka1:8003/eureka/
#启动重试机制
spring.cloud.loadbalancer.retry.enabled=true 
##断路器
hystrix.command.default.execution.isolation.thread.timeoutInMilliseconds=14000
##连接超时时间
#client-service.ribbon.ConnectTimeout=1000
##处理超时时间
#client-service.ribbon.ReadTimeout=1000

req.fact.connect-timeout=1000 
req.fact.connection-request-timeout=1000
req.fact.read-timeout=3000
#是否对所有请求都进行重试
client-service.ribbon.OKToRetryOnAllOperations=true
#重试切换实例得次数
client-service.ribbon.MaxAutoRetriesNextServer=1
#重试切次数
client-service.ribbon.MaxAutoRetries=5

• 注意：

client-service.ribbon.ConnectTimeout=1000
client-service.ribbon.ReadTimeout=1000
这样配置没起作用  建议使用RestTemplate原始配置
hystrix.command.default.execution.isolation.thread.timeoutInMilliseconds：断路器得超时时间需要大于ribbon得超时时间，不然不会发生重试。

• RetryApplication类

@EnableCircuitBreaker //开启断路器
@EnableRetry    //启动Retry框架重试机制
@EnableDiscoveryClient //我是服务 注册到服务中心上
@SpringBootApplication
public class RetryApplication {


    @Bean
    @ConfigurationProperties(prefix = "req.fact")
    public HttpComponentsClientHttpRequestFactory httpComponentsClientHttpRequestFactory(){
        return new HttpComponentsClientHttpRequestFactory();
    }

    @Bean
    @LoadBalanced
    public RestTemplate restTemplate(){
        return new RestTemplate(httpComponentsClientHttpRequestFactory());
    }
    public static void main(String[] args) {
        SpringApplication.run(RetryApplication.class, args);
    }
}

• IndexController类

@RestController
public class IndexController {

    @Autowired
    RestTemplate restTemplate;

    @RequestMapping(value = "/retry", method = {RequestMethod.GET})
    @HystrixCommand(fallbackMethod = "retryFailback")
    public String retry(){
        ResponseEntity<String> response = restTemplate.getForEntity("http://client-service/retry", String.class);
        System.err.println("body:" + response.getBody());
        return "----- retry ok -----";
    }

    public String retryFailback(){
        System.err.println("-----启动降级策略----");
        return "-----启动降级策略----";
    }
}

• 在spring-cloud-02-ribbon-client1项目得Controller中添加

 @RequestMapping(value = "/retry", method = {RequestMethod.GET})
    public String retry() {
        System.err.println("client1  1 call-------------");
        try {
            Thread.sleep(4000);
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
        return "client 1";
    }

• 在spring-cloud-02-ribbon-client2项目得Controller中添加

 @RequestMapping(value = "/retry", method = {RequestMethod.GET})
    public String retry() {
        System.err.println("client1  2 call-------------");
        return "client 2";
    }

• 在启动http://localhost:1006/retry时，通过Ribbon负载均衡会访问client1或者client2项目得retry方法，当访问到client1得retry方法时，配置得读取超时时间为3ms，重试次数为5次，retry方法睡眠为4ms，就是每次访问retry方法时重试，当重试次数6次后，会采用降级策略，因为熔断器配置得为14ms，超过熔断器得时间就会采用降级策略。
• GitHub地址
• hystrix 先入个门