How to Integrate Zipkin with Spring Cloud Sleuth for Distributed Tracing

Building on a previous article about using ELK for log collection, this tutorial explains why traditional log analysis falls short for latency monitoring and introduces Zipkin, an open‑source distributed tracing system based on Google Dapper.

Zipkin Overview

Zipkin collects trace spans from services, stores them, and provides a REST API and a Web UI for querying trace timelines and service dependency graphs.

Step 1 – Build a Zipkin Server

Create a Spring Boot project named zipkin-server and add the following dependencies to pom.xml:

<parent>
  <groupId>org.springframework.boot</groupId>
  <artifactId>spring-boot-starter-parent</artifactId>
  <version>1.3.7.RELEASE</version>
</parent>

<dependencies>
  <dependency>
    <groupId>io.zipkin.java</groupId>
    <artifactId>zipkin-server</artifactId>
  </dependency>
  <dependency>
    <groupId>io.zipkin.java</groupId>
    <artifactId>zipkin-autoconfigure-ui</artifactId>
  </dependency>
</dependencies>

<dependencyManagement>
  <dependencies>
    <dependency>
      <groupId>org.springframework.cloud</groupId>
      <artifactId>spring-cloud-dependencies</artifactId>
      <version>Brixton.SR5</version>
      <type>pom</type>
      <scope>import</scope>
    </dependency>
  </dependencies>
</dependencyManagement>

Add a main class annotated with @EnableZipkinServer and @SpringBootApplication:

@EnableZipkinServer
@SpringBootApplication
public class ZipkinApplication {
    public static void main(String[] args) {
        SpringApplication.run(ZipkinApplication.class, args);
    }
}

Configure the server port (default 9411) in application.properties:

spring.application.name=zipkin-server
server.port=9411

Run the application and open http://localhost:9411/ to see the Zipkin UI.

Step 2 – Configure Spring Cloud Sleuth to Send Traces to Zipkin

In each microservice (e.g., trace-1 and trace-2) add the Zipkin starter dependency:

<dependency>
  <groupId>org.springframework.cloud</groupId>
  <artifactId>spring-cloud-sleuth-zipkin</artifactId>
</dependency>

Set the Zipkin server address in application.properties (the HTTP URL can be omitted if the server runs on the default port 9411):

spring.zipkin.base-url=http://localhost:9411

Collecting Traces via Message Middleware

To send traces asynchronously through RabbitMQ, add the stream extensions:

<dependency>
  <groupId>org.springframework.cloud</groupId>
  <artifactId>spring-cloud-sleuth-stream</artifactId>
</dependency>

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

Remove the HTTP‑only spring.zipkin.base-url property and add RabbitMQ settings:

spring.rabbitmq.host=localhost
spring.rabbitmq.port=5672
spring.rabbitmq.username=springcloud
spring.rabbitmq.password=123456

On the Zipkin server side, add the corresponding collector dependencies:

<dependency>
  <groupId>org.springframework.cloud</groupId>
  <artifactId>spring-cloud-sleuth-zipkin-stream</artifactId>
</dependency>

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

After rebuilding and restarting both the client services and the Zipkin server, a RabbitMQ exchange named sleuth will receive trace messages.

Testing and Analysis

Start eureka-server, trace-1, trace-2, and the Zipkin server. Send a few HTTP requests to http://localhost:9101/trace-1. When the trace is sampled (e.g., using AlwaysSampler), the trace appears in the Zipkin UI, and the corresponding RabbitMQ message can be seen in the sleuth exchange.

In the Zipkin UI you can explore individual trace details, view the request latency breakdown, and inspect the automatically generated service dependency graph:

Complete Example

The full source code is available in two repositories:

GitHub: https://github.com/dyc87112/SpringCloud-Learning/

Gitee: https://gitee.com/didispace/SpringCloud-Learning/

These projects contain the trace-1, trace-2, and zipkin-server modules demonstrated above.