众所周知,Spring Cloud Sleuth有两种方式整合Zipkin:
HTTP直连Zipkin方式
MQ方式,架构图如下:
Spring Cloud Edgware及更高版本中,Sleuth使用MQ方式整合Zipkin的玩法发生了巨大改变。本文将贴出新旧版本中Sleuth如何整合Zipkin的具体操作。MQ使用的是RabbitMQ(读者也可使用Kafka)。
Dalston及更低版本
如果您使用Spring Cloud Dalston或更低版本,那么整合步骤如下:
服务器端
-
依赖:
<dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-sleuth-zipkin-stream</artifactId> </dependency> <dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-starter-sleuth</artifactId> </dependency> <dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-stream-binder-rabbit</artifactId> </dependency> <dependency> <groupId>io.zipkin.java</groupId> <artifactId>zipkin-autoconfigure-ui</artifactId> </dependency>
启动类上添加注解
@EnableZipkinStreamServer
。-
配置:
server: port: 9411 spring: rabbitmq: host: localhost port: 5672 username: guest password: guest
微服务端
-
加依赖:
<dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-sleuth-stream</artifactId> </dependency> <dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-starter-sleuth</artifactId> </dependency> <dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-stream-binder-rabbit</artifactId> </dependency>
-
配置:
spring: rabbitmq: host: localhost port: 5672 username: guest password: guest
Edgware及更高版本
对于Edgware,以上方式也可使用,但目前已标注为废弃
,未来将被删除!
从Edgware开始,应如下整合Zipkin——
服务器端
-
依赖:
<dependency> <groupId>io.zipkin.java</groupId> <artifactId>zipkin-autoconfigure-ui</artifactId> </dependency> <dependency> <groupId>io.zipkin.java</groupId> <artifactId>zipkin-server</artifactId> </dependency> <!-- 使用消息的方式收集数据(使用rabbitmq) --> <dependency> <groupId>io.zipkin.java</groupId> <artifactId>zipkin-autoconfigure-collector-rabbitmq</artifactId> <version>2.3.1</version> </dependency>
启动类上添加
@EnableZipkinServer
。-
配置:
server: port: 9411 zipkin: collector: rabbitmq: addresses: localhost:5672 password: guest username: guest queue: zipkin
微服务端
-
依赖:
<dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-starter-zipkin</artifactId> </dependency> <dependency> <groupId>org.springframework.amqp</groupId> <artifactId>spring-rabbit</artifactId> </dependency>
-
配置:
spring: rabbitmq: host: localhost port: 5672 username: guest password: guest zipkin: rabbitmq: queue: zipkin
对比及分析
对比后不难发现,从Edgware开始,Sleuth基于MQ整合Zipkin更加的简化!不过,改动还是比较大的。官方为什么要做这个改进呢?答案如下——
- Spring Cloud Edgware之前的版本使用
Zipkin 1.x
,要想MQ方式收集数据,需整合spring-cloud-sleuth-stream
。而在Edgware及更高版本中,使用Zipkin 2.x
。Zipkin 2.x
本身已支持基于MQ的数据收集方式,故而spring-cloud-sleuth-stream
将被废弃! - 两种使用方式不兼容,请读者务必注意!