Spring Cloud Stream使用

╰+哭是因爲堅強的太久メ 2022-08-31 06:36 27阅读 0赞

Spring Cloud Stream对Spring Cloud体系中的Mq进⾏了很好的上层抽象，可以让我们与具体消息中间件解耦合，屏蔽掉了底层具体MQ消息中间件的细节差异，就像Hibernate屏蔽掉了具体数据库（Mysql/Oracle⼀样）。如此⼀来，我们学习、开发、维护MQ都会变得轻松。⽬前Spring Cloud Stream原生⽀持RabbitMQ和Kafka，阿里在这个基础上提供了RocketMQ的支持

## 简单使用Spring Cloud Stream 构建基于RocketMQ的生产者和消费者 ##

### 生产者 ###

pom文件中加入依赖

<dependencies>
            <dependency>
                <groupId>org.springframework.cloud</groupId>
                <artifactId>spring-cloud-starter-netflix-eureka-client</artifactId>
            </dependency>
            <dependency>
                <groupId>com.alibaba.cloud</groupId>
                <artifactId>spring-cloud-starter-stream-rocketmq</artifactId>
                <version>2.1.0.RELEASE</version>
            </dependency>
    
        </dependencies>

配置文件中增加关于Spring Cloud Stream binder和bindings的配置

spring:
      application:
        name: zhao-cloud-stream-producer
      cloud:
        stream:
          rocketmq:
            binder:
              name-server: 127.0.0.1:9876
            bindings:
              output:
                producer:
                  group: test
                  sync: true
          bindings:
            output:
              destination: stream-test-topic
              content-type: text/plain # 内容格式。这里使用 JSON

其中destination代表生产的数据发送到的topic 然后定义一个channel用于数据发送

import org.springframework.cloud.stream.annotation.Output;
    import org.springframework.messaging.MessageChannel;
    
    public interface TestChannel {
        @Output("output")
        MessageChannel output();
    }

最后构造数据发送的接口

@Controller
    public class SendMessageController {
        @Resource
        private TestChannel testChannel;
    
        @ResponseBody
        @RequestMapping(value = "send", method = RequestMethod.GET)
        public String sendMessage() {
            String messageId = UUID.randomUUID().toString();
            Message<String> message = MessageBuilder
                    .withPayload("this is a test:" + messageId)
                    .setHeader(MessageConst.PROPERTY_TAGS, "test")
                    .build();
            try {
                testChannel.output().send(message);
                return messageId + "发送成功";
            } catch (Exception e) {
                return messageId + "发送失败，原因：" + e.getMessage();
            }
        }
    }

### 消费者 ###

消费者的pom引入与生产者相同，在此不再赘述，配置时需要将stream的output修改为input并修改对应属性

spring:
      application:
        name: zhao-cloud-stream-consumer
      cloud:
        stream:
          rocketmq:
            binder:
              name-server: 127.0.0.1:9876
            bindings:
              input:
                consumer:
                  tags: test
          bindings:
            input:
              destination: stream-test-topic
              content-type: text/plain # 内容格式。这里使用 JSON
              group: test

另外关于channel的构造也要做同样的修改

import org.springframework.cloud.stream.annotation.Input;
    import org.springframework.messaging.SubscribableChannel;
    
    public interface TestChannel {
        @Input("input")
        SubscribableChannel input();
    }

最后我在启动类中对收到的消息进行了监听

@StreamListener("input")
        public void receiveInput(@Payload Message message) throws ValidationException {
            System.out.println("input1 receive: " + message.getPayload() + ", foo header: " + message.getHeaders().get("foo"));
        }

测试结果 ![file][]

![file][file 1]

## Stream其他特性 ##

### 消息发送失败的处理 ###

消息发送失败后悔发送到默认的一个“topic.errors"的channel中(topic是配置的destination)。要配置消息发送失败的处理，需要将错误消息的channel打开 消费者配置如下

在启动类中配置错误消息的Channel信息

@Bean("stream-test-topic.errors")
        MessageChannel testoutPutErrorChannel(){
            return new PublishSubscribeChannel();
        }

新建异常处理service

import org.springframework.integration.annotation.ServiceActivator;
    import org.springframework.messaging.Message;
    import org.springframework.stereotype.Service;
    
    @Service
    public class ErrorProducerService {
    
        @ServiceActivator(inputChannel = "stream-test-topic.errors")
        public void receiveProducerError(Message message){
            System.out.println("receive error msg :"+message);
        }
    }

当发生异常时，由于测试类中已经将异常捕获，处理发送异常主要是在这里进行。模拟，应用与rocketMq断开的场景。可见 ![file][file 2] ![file][file 3]

### 消费者错误处理 ###

首先增加配置为

增加相应的模拟异常的操作

@StreamListener("input")
        public void receiveInput(@Payload Message message) throws ValidationException {
            //System.out.println("input1 receive: " + message.getPayload() + ", foo header: " + message.getHeaders().get("foo"));
            throw new RuntimeException("oops");
        }
        @ServiceActivator(inputChannel = "stream-test-topic.test.errors")
        public void receiveConsumeError(Message message){
            System.out.println("receive error msg"+message.getPayload());
        }

![file][file 4]

代码地址[https://github.com/zhendiao/deme-code/tree/main/zp][https_github.com_zhendiao_deme-code_tree_main_zp]

> 欢迎搜索关注本人的公众号【微瞰技术】,以及总结的分类面试题[https://github.com/zhendiao/JavaInterview][https_github.com_zhendiao_JavaInterview]

![file][file 5]

[file]: /images/20220829/62daf981c7a2417fbd77ddf999aa2fd9.png
[file 1]: /images/20220829/c5b4606984084d7bbcfc2501cfdd3185.png
[file 2]: https://img-blog.csdnimg.cn/20210723123416833.png
[file 3]: /images/20220829/dfa4d80e64234335bf0284eb35f21c37.png
[file 4]: /images/20220829/eb47897be1004103af6e54680ac765a9.png
[https_github.com_zhendiao_deme-code_tree_main_zp]: https://github.com/zhendiao/deme-code/tree/main/zp
[https_github.com_zhendiao_JavaInterview]: https://github.com/zhendiao/JavaInterview
[file 5]: /images/20220829/a04a343d0f8c4c689a2a15cd75261a60.png