Master Distributed Tracing with Spring Cloud Sleuth and Zipkin: A Step‑by‑Step Guide

Spring Cloud Sleuth Overview

As systems grow, service‑to‑service calls become complex, making it hard to pinpoint delays or errors. Spring Cloud Sleuth provides distributed tracing to visualize the call chain of a request across multiple services.

Adding Request Tracing to Services

We demonstrate tracing between

user-service

and

ribbon-service

. The

ribbon-service

calls

user-service

via

RestTemplate

.

Add tracing support to both services.

Include the required dependencies:

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

Configure

application.yml

to point to the Zipkin server and set the sampling probability:

spring:
  zipkin:
    base-url: http://localhost:9411
  sleuth:
    sampler:
      probability: 0.1  # set Sleuth sampling rate

Integrating Zipkin for Log Collection and Analysis

Zipkin, an open‑source project from Twitter, collects and visualizes tracing data generated by Spring Cloud Sleuth.

For Spring Boot 2.0+, you can run the Zipkin server without building it yourself:

java -jar zipkin-server-2.12.9-exec.jar

Access the Zipkin UI at

http://localhost:9411

.

Start

eureka-server

,

ribbon-service

, and

user-service

, then repeatedly call

http://localhost:8301/user/1

. Zipkin will display the traced requests, showing each service’s latency.

Persisting Trace Information

By default Zipkin stores traces in memory, which are lost on restart. To persist data, you can configure Zipkin to use Elasticsearch.

Install Elasticsearch

Download Elasticsearch 6.2.2 zip package and unzip it.

Run

bin/elasticsearch.bat

to start Elasticsearch.

Configure Zipkin to Store Traces in Elasticsearch

Run Zipkin with storage parameters:

# STORAGE_TYPE: storage type, ES_HOSTS: Elasticsearch address
java -jar zipkin-server-2.12.9-exec.jar --STORAGE_TYPE=elasticsearch --ES_HOSTS=localhost:9200

Restart

user-service

and

ribbon-service

, then invoke the endpoint again. If Kibana is installed, you can view the stored traces.

Additional Startup Parameters

Refer to the official Zipkin documentation for more options: https://github.com/openzipkin/zipkin/tree/master/zipkin-server#elasticsearch-storage

Modules Used in the Project

springcloud-learning
├── eureka-server   # Eureka registry
├── user-service    # Provides CRUD API for User objects
└── ribbon-service  # Service used for Ribbon call testing

Project Source Code

https://github.com/macrozheng/springcloud-learning