700字范文 > 【Spring Cloud Alibaba 温故而知新】（五）SpringCloud Sleuth + Zipkin：分布式日志追踪

【Spring Cloud Alibaba 温故而知新】（五）SpringCloud Sleuth + Zipkin：分布式日志追踪

时间：2020-05-25 04:26:10

SpringCloud Sleuth 必知必会 SpringCloud Sleuth 实现的功能是：它会自动为当前应用构建起各通信通道的跟踪机制通过诸如 RabbitMQ、Kafka（或者其他任何 SpringCloud Sleuth 绑定器实现的消息中间件）传递的请求通过 Zuul、Gateway 代理传递的请求通过 RestTemplate 发起的请求 SpringCloud Sleuth 跟踪实现原理为了实现请求跟踪：当请求发送到分布式系统的入口端点时，只需要服务跟踪框架为该请求创建一个唯一的跟踪标识 Trace ID为了统计各处理单元的时间延迟，当请求到达各个服务组件时，或是处理逻辑到达某个状态时，也通过一个唯一标识来标记它的开始、具体过程以及结束，Span ID

Span ID 如果要计算时间延迟，可以通过 spanIdA11 - spanIdA1

8.1.2 Zipkin 是什么

Zipkin 的基础概念 Zipkin 解决微服务架构中的延迟问题，包括数据的收集、存储、查找和展现Zipkin 有四大核心组件构成 Collector：收集器组件Storge：存储组件API：RESTFul API，提供外部访问接口UI：Web UI，提供可视化查询页面

8.2.1 集成 SpringCloud Sleuth 实现微服务通信跟踪

8.2.1.1 集成步骤

保证服务与服务之间存在跨进程通信Maven 依赖

8.2.1.2 编写测试代码

sca-commerce-gateway 与 sca-commerce-alibaba-nacos-client 添加 Maven 依赖

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

sca-commerce-alibaba-nacos-client 编写测试业务代码与控制层代码

SleuthTraceInfoService

package merce.service;import brave.Tracer;import lombok.RequiredArgsConstructor;import lombok.extern.slf4j.Slf4j;import org.springframework.stereotype.Service;/*** @author eddie.lee* @blog * @description 使用代码更直观的看到 Sleuth 生成的相关跟踪信息*/@Slf4j@Service@RequiredArgsConstructorpublic class SleuthTraceInfoService {/** brave.Tracer 跟踪对象 */private final Tracer tracer;/*** 打印当前的跟踪信息到日志中*/public void logCurrentTraceInfo() {log.info("Sleuth trace id: [{}]", tracer.currentSpan().context().traceId());log.info("Sleuth span id: [{}]", tracer.currentSpan().context().spanId());}}

SleuthTraceInfoController

package merce.controller;import merce.service.SleuthTraceInfoService;import lombok.RequiredArgsConstructor;import lombok.extern.slf4j.Slf4j;import org.springframework.web.bind.annotation.GetMapping;import org.springframework.web.bind.annotation.RequestMapping;import org.springframework.web.bind.annotation.RestController;/*** @author eddie.lee* @blog * @description 打印跟踪信息*/@Slf4j@RestController@RequestMapping("/sleuth")@RequiredArgsConstructorpublic class SleuthTraceInfoController {private final SleuthTraceInfoService traceInfoService;/*** 打印日志跟踪信息*/@GetMapping("/trace-info")public void logCurrentTraceInfo() {traceInfoService.logCurrentTraceInfo();}}

8.2.1.2 测试请求与查看控制台日志

发起请求

### 查看 Sleuth 跟踪信息GET http://127.0.0.1:9001/edcode/scacommerce-nacos-client/sleuth/trace-infoAccept: application/jsonsca-commerce-user: eyJhbGciOiJSUzI1NiJ9.eyJzY2EtY29tbWVyY2UtdXNlciI6IntcImlkXCI6MTEsXCJ1c2VybmFtZVwiOlwiZWRkaWVAcXEuY29tXCJ9IiwianRpIjoiZjQ3M2NhZjctY2RjMi00ZmE4LWExNzQtZjZhYmQ5ZDFjMzAzIiwiZXhwIjoxNjM1ODY4ODAwfQ.iTtQE2gHzjPxVP5SEFHrDBkvrzI-yt6oy-w1x--Q3ahhTvYLTiYnvndtIx7IIyYipr_ayZnAQyluPt3oiLaS80r9qByaN3zQF-6gBW_wu_fd0yd89hIjPnQeP1mY2NcchV2FaMUW7Jlq8CUDPurEhW4GUDXOqBXgmxai5UTu4yoXBUfyXUXznKTx697cGo5aoVKTAKvMReJg-77n5sQuafZNDu6pz2D1KMvEucNyZtbXw0JRIl1CsK777Jt3IG1bnOnwRBt8o1tkodZ3zJbfgTGVCHJmfEuUnXwdf4DLAq568pNVvylPLh4_r-UUGGxE6Az9XwOtl1w4vzK1M2ATzwtoken: edcode

响应信息

GET http://127.0.0.1:9001/edcode/scacommerce-nacos-client/sleuth/trace-infoHTTP/1.1 200 OKtransfer-encoding: chunkedContent-Type: application/jsonDate: Tue, 02 Nov 13:04:55 GMT{"code": 0,"message": "","data": null}Response code: 200 (OK); Time: 1160ms; Content length: 35 bytes

查看日志

sca-commerce-gateway

-11-02 21:04:55.332 INFO [sca-commerce-gateway,353ea734cc43d6ee,353ea734cc43d6ee,true] 1060 --- [ctor-http-nio-2] flix.config.ChainedDynamicProperty : Flipping property: sca-commerce-nacos-client.ribbon.ActiveConnectionsLimit to use NEXT property: niws.loadbalancer.availabilityFilteringRule.activeConnectionsLimit = 2147483647-11-02 21:04:55.347 INFO [sca-commerce-gateway,353ea734cc43d6ee,353ea734cc43d6ee,true] 1060 --- [ctor-http-nio-2] flix.loadbalancer.BaseLoadBalancer : Client: sca-commerce-nacos-client instantiated a LoadBalancer: DynamicServerListLoadBalancer:{NFLoadBalancer:name=sca-commerce-nacos-client,current list of Servers=[],Load balancer stats=Zone stats: {},Server stats: []}ServerList:null-11-02 21:04:55.353 INFO [sca-commerce-gateway,353ea734cc43d6ee,353ea734cc43d6ee,true] 1060 --- [ctor-http-nio-2] c.n.l.DynamicServerListLoadBalancer: Using serverListUpdater PollingServerListUpdater-11-02 21:04:55.372 INFO [sca-commerce-gateway,353ea734cc43d6ee,353ea734cc43d6ee,true] 1060 --- [ctor-http-nio-2] flix.config.ChainedDynamicProperty : Flipping property: sca-commerce-nacos-client.ribbon.ActiveConnectionsLimit to use NEXT property: niws.loadbalancer.availabilityFilteringRule.activeConnectionsLimit = 2147483647-11-02 21:04:55.374 INFO [sca-commerce-gateway,353ea734cc43d6ee,353ea734cc43d6ee,true] 1060 --- [ctor-http-nio-2] c.n.l.DynamicServerListLoadBalancer: DynamicServerListLoadBalancer for client sca-commerce-nacos-client initialized: DynamicServerListLoadBalancer:{NFLoadBalancer:name=sca-commerce-nacos-client,current list of Servers=[192.168.3.192:8000],Load balancer stats=Zone stats: {unknown=[Zone:unknown;Instance count:1;Active connections count: 0;Circuit breaker tripped count: 0;Active connections per server: 0.0;]},Server stats: [[Server:192.168.3.192:8000;Zone:UNKNOWN;Total Requests:0;Successive connection failure:0;Total blackout seconds:0;Last connection made:Thu Jan 01 08:00:00 CST 1970;First connection made: Thu Jan 01 08:00:00 CST 1970;Active Connections:0;total failure count in last (1000) msecs:0;average resp time:0.0;90 percentile resp time:0.0;95 percentile resp time:0.0;min resp time:0.0;max resp time:0.0;stddev resp time:0.0]]}ServerList:com.alibaba.cloud.nacos.ribbon.NacosServerList@72186c8f-11-02 21:04:55.592 INFO [sca-commerce-gateway,353ea734cc43d6ee,353ea734cc43d6ee,true] 1060 --- [ctor-http-nio-2] c.e.c.filter.GlobalElapsedLogFilter: [/edcode/scacommerce-nacos-client/sleuth/trace-info] elapsed: [1034ms]-11-02 21:04:56.358 INFO [sca-commerce-gateway,,,] 1060 --- [erListUpdater-0] flix.config.ChainedDynamicProperty : Flipping property: sca-commerce-nacos-client.ribbon.ActiveConnectionsLimit to use NEXT property: niws.loadbalancer.availabilityFilteringRule.activeConnectionsLimit = 2147483647

sca-commerce-alibaba-nacos-client

-11-02 21:04:55.543 INFO [sca-commerce-nacos-client,353ea734cc43d6ee,c85be2c1bb127558,true] 33740 --- [nio-8000-exec-1] c.e.c.service.SleuthTraceInfoService: Sleuth trace id: [3836687777773377262]-11-02 21:04:55.543 INFO [sca-commerce-nacos-client,353ea734cc43d6ee,c85be2c1bb127558,true] 33740 --- [nio-8000-exec-1] c.e.c.service.SleuthTraceInfoService: Sleuth span id: [-4009361721548180136]

解析：[sca-commerce-nacos-client,353ea734cc43d6ee,c85be2c1bb127558,true]
第一行：service name
第二行：trace id
第三行：span id

8.3.1 搭建 Zipkin Server 实现对跟踪信息的收集

8.3.1.1 ZS搭建步骤

Tips：SpringCloud Finchley 版本（包含）之后，官方不建议自己搭建 Zipkin-Server，提供了已经打包好的jar文件（SpringBoot工程），直接下载启动即可下载地址 curl -sSL https://zipkin.io/quickstart.sh | bash -s 选择自己需要的版本即可选择 *.exec.jar 结尾的 jar

8.3.1.2 Linux 终端

[root@localhost opt]# curl -sSL https://zipkin.io/quickstart.sh | bash -s Thank you for trying Zipkin!This installer is provided as a quick-start helper, so you can try Zipkin outwithout a lengthy installation process.Fetching version number of latest io.zipkin:zipkin-server release...Latest release of io.zipkin:zipkin-server seems to be 2.23.4Downloading io.zipkin:zipkin-server:2.23.4:exec to zipkin.jar...> curl -fL -o 'zipkin.jar' '/maven2/io/zipkin/zipkin-server/2.23.4/zipkin-server-2.23.4-exec.jar'% Total % Received % Xferd Average Speed Time TimeTime CurrentDload Upload Total Spent Left Speed100 59.0M 100 59.0M 00 341460 0:30:14 0:30:14 --:--:-- 33309Verifying checksum...> curl -fL -o 'zipkin.jar.md5' '/maven2/io/zipkin/zipkin-server/2.23.4/zipkin-server-2.23.4-exec.jar.md5'% Total % Received % Xferd Average Speed Time TimeTime CurrentDload Upload Total Spent Left Speed100 32 100 32 00170 0:00:01 0:00:01 --:--:-- 17> md5sum -c <<< "$(cat zipkin.jar.md5) zipkin.jar"zipkin.jar: OKChecksum for zipkin.jar passes verificationVerifying GPG signature of zipkin.jar...> curl -fL -o 'zipkin.jar.asc' '/maven2/io/zipkin/zipkin-server/2.23.4/zipkin-server-2.23.4-exec.jar.asc'% Total % Received % Xferd Average Speed Time TimeTime CurrentDload Upload Total Spent Left Speed100 833 100 833 00910 0:00:09 0:00:09 --:--:-- 180GPG signing key is not known, skipping signature verification.Use the following commands to manually verify the signature of zipkin.jar:gpg --keyserver --recv FF31B515# Optionally trust the key via 'gpg --edit-key FF31B515', then typing 'trust',# choosing a trust level, and exiting the interactive GPG session by 'quit'gpg --verify zipkin.jar.asc zipkin.jarYou can now run the downloaded executable jar:java -jar zipkin.jar[root@localhost opt]# nohup java -jar zipkin.jar &[1] 30238[root@localhost opt]# nohup: ignoring input and appending output to ‘nohup.out’

8.3.1.3 Zipkin Web UI

访问：IP:9411

8.3.2 配置 Zipkin Server 实现对跟踪信息的收集

配置 ZS 为什么需要对 ZS 做自定义配置？默认情况下， ZS 将跟踪信息存储在内存中（JVM），重启会丢失ZS 默认使用 HTTP 方式上报跟踪数据，性能较差 ZS 配置 MySQL 跟踪数据持久化（同时它也支持 ES） MySQL 中添加数据表：/openzipkin/zipkin/blob/master/zipkin-storage/mysql-v1/src/main/resources/mysql.sqlZS 启动指定 MySQL 路径

8.4.1 SpringCloud Sleuth 整合 Zipkin 实现分布式链路跟踪、收集

SpringCloud Sleuth 整合 Zipkin 步骤

简单的两个步骤（Zipkin Server 使用 MySQL 实现跟踪数据持久化） Maven 依赖bootstrap.yml 中增加 Zipkin 的配置

下载、安装 Kafka

下载 Kafka：/quickstart解压、启动 ZK 和 Kafka Server 即可（使用默认配置）

8.4.1.1 下载与解压 Kafka

Downloads

/downloads

Linux 步骤

[root@localhost opt]# wget /kafka/3.0.0/kafka_2.13-3.0.0.tgz[root@localhost opt]# tar -zxf kafka_2.13-3.0.0.tgz [root@localhost opt]# ls -la | grep kafkadrwxr-xr-x. 8 root root 134 Nov 3 01:48 kafka_2.13-3.0.0-rw-r--r--. 1 root root 86396520 Sep 20 04:46 kafka_2.13-3.0.0.tgz

8.4.1.2 启动 zookeeper 与 Kafka

对外访问需要修改 Kafka 的 server.properties

找到 advertised.listeners 或者自行添加 advertised.listeners 指定 IP 地址

[root@localhost kafka_2.13-3.0.0]# vim /opt/kafka_2.13-3.0.0/config/server.propertiesadvertised.listeners=PLAINTEXT://192.168.3.250:9092

后台启动 zookeeper

[root@localhost kafka_2.13-3.0.0]# nohup /opt/kafka_2.13-3.0.0/bin/zookeeper-server-start.sh config/zookeeper.properties &[1] 31998[root@localhost kafka_2.13-3.0.0]# nohup: ignoring input and appending output to ‘nohup.out’

后台启动 kafka

[root@localhost kafka_2.13-3.0.0]# nohup /opt/kafka_2.13-3.0.0/bin/kafka-server-start.sh config/server.properties &[1] 32574[root@localhost kafka_2.13-3.0.0]# nohup: ignoring input and appending output to ‘nohup.out’

8.4.1.3 运行 ZipKin 关联 Kafka 与 MySQL

[root@localhost opt]# nohup java -DKAFKA_BOOTSTRAP_SERVERS=127.0.0.1:9092 -jar zipkin.jar --STORAGE_TYPE=mysql --MYSQL_USER=root --MYSQL_PASS=123456 --MYSQL_HOST=127.0.0.1 --MYSQL_TCP_PORT=3306 --MYSQL_DB=zipkin &[1] 601[root@localhost opt]# nohup: ignoring input and appending output to ‘nohup.out’

连接kafka：-DKAFKA_BOOTSTRAP_SERVERS=127.0.0.1:9092
连接MySQL：–STORAGE_TYPE=mysql --MYSQL_USER=root --MYSQL_PASS=123456 --MYSQL_HOST=127.0.0.1 --MYSQL_TCP_PORT=3306 --MYSQL_DB=zipkin

8.4.1.4 检查 Linux 服务是否启动

8.4.1.4 IDEA 启动服务与测试发起请求

Maven 依赖（zipkin 与 kafka）

<dependency><groupId>org.springframework.cloud</groupId><artifactId>spring-cloud-starter-zipkin</artifactId></dependency><dependency><groupId>org.springframework.kafka</groupId><artifactId>spring-kafka</artifactId><version>2.5.0.RELEASE</version></dependency>

修改 Gateway 与 Nacos-Client 服务的配置文件

spring:kafka:bootstrap-servers: ${KAFKA_SERVER:127.0.0.1}:${KAFKA_PORT:9092}producer:retries: 3consumer:auto-offset-reset: latestzipkin:sender:type: ${ZIPKIN_KAFKA_SENDER:web} # 默认是 webbase-url: http://${ZIPKIN_URL:localhost}:${ZIPKIN_PORT:9411}/

两个服务都需要添加 kafka 与 zipkin 的连接信息

启动服务

启动以下服务 AuthorityCenterApplication :7000/ # 如果 token 过期，需要重新签发NacosClientApplication :8000/GatewayApplication :9001/

请求测试

sca-commerce-gateway/src/main/resources/http/nacos-client.http

### 查询服务GET http://127.0.0.1:9001/edcode/scacommerce-nacos-client/nacos-client/service-instance?serviceId=sca-commerce-gatewayAccept: application/jsonsca-commerce-user: eyJhbGciOiJSUzI1NiJ9.eyJzY2EtY29tbWVyY2UtdXNlciI6IntcImlkXCI6MTEsXCJ1c2VybmFtZVwiOlwiZWRkaWVAcXEuY29tXCJ9IiwianRpIjoiMWU1MGI2ZWYtNmUzOS00YmY2LWJlMjktZDc4NWU3NWQyNmY1IiwiZXhwIjoxNjM1OTU1MjAwfQ.P7GxZuMUrgiMUbD4dNYzQiV3A6YkaFpvlzg8cpBdu_hvxqDsVEuuiYODQSzZPQeN3xTQPbJ70zkSY084HV7Vsk929en1lqNiX_dpQEuGSbz2JSPqyJuLZ6v7hRX9GI32sPrZAnaKVXMdeHUXCMMmaS1L3osimSvAlaoDE0n2UukDLgu83xRlL3bddHIJbmFD5BrV6Y-u9d-blqXPOpxFEYkdwS_XrljYiULTH7Olr71TAwODUPdttnmVhHPXB0_dnOG5DZMOC0OxqokHGZJ7CC86paE4TvdNPwqotB6u6zh_d_YCCBWM3t1LmKYB6E_bnz2taL5Q4AYHlRaZZotaAAtoken: edcode ## HeaderTokenGatewayFilter###

8.4.1.5 Zipkin Web UI

如何简单的使用

打开 http://192.168.3.250:9411

默认 all，然后直接查找，会显示所有的请求信息，点击其中一条

就会看到该请求的所有经过哪些服务，耗时多少

同样，也可以在终端拿 trace Id：669b59f38adf2c38 去跟踪链路

trace Id 搜索框：669b59f38adf2c38

如何查看服务之间的依赖关系

点击上方的【依赖】

8.5.1 Spring Cloud Sleuth 设置采样率、抽样收集策略

8.5.1.1 Spring Cloud Sleuth 采样收集

收集跟踪信息是一把双刃剑，需要做好权衡收集的跟踪信息越多，越能反映出系统的实际运行情况高并发场景下，大量的请求调用会产生海量的跟踪日志信息，性能开销太大

开发与测试环境可以使用高的采样率，但是生产环境建议不要这么做。

可以自由选择 Zipkin brave 自带的两个抽样策略 ProbabilityBasedSampler采样率策略默认使用的策略，以请求百分比的方式配置和手机跟踪信息：它的默认值为 0.1，代表手机 10% 的请求跟踪信息spring.sleuth.sampler.probability=0.5 RateLimitingSampler抽样策略限速采集，也就是说它可以用来限制每秒追踪请求的最大数量，优先级更高spring.sleuth.sampler.rate=10 ## 一秒最大只有10个跟踪策略给采集

8.5.1.2 bootstrap.yml 配置 Sleuth

sca-commerce-alibaba-nacos-client

spring:sleuth:sampler:# RateLimitingSampler 抽样策略，设置了限速采样，spring.sleuth.sampler.probability 属性值无效rate: 100 # 每秒间隔接受的 trace 量# Probability 抽样策略probability: 1.0 # 采样比例，1.0 表示 100%, 默认：0.1

8.5.1.3 代码配置 Sleuth

package merce.sampler;import brave.sampler.RateLimitingSampler;import brave.sampler.Sampler;import org.springframework.context.annotation.Bean;import org.springframework.context.annotation.Configuration;/*** @author eddie.lee* @blog * @description 使用配置的方式设定抽样率（二选一）*/@Configurationpublic class SamplerConfig {/*** 限速采集（推荐）*/@Beanpublic Sampler sampler() {return RateLimitingSampler.create(100);}// /**//* 概率采集, 默认的采样策略, 默认值是 0.1//*/// @Bean// public Sampler defaultSampler() {// return ProbabilityBasedSampler.create(0.5f);// }}

代码与Yaml文件配置是二选一，方便简洁肯定是 Yaml配置
限速采集与概率采集无论是代码还是 Yaml文件都是二选一

8.6.1 SpringCloud Sleuth+Zipkin 分布式日志追踪总结

8.6.1.1 SpringCloud Sleuth+Zipkin 逻辑架构图

跟踪、收集所涉及的三个组件（模块）Sleuth、Zipkin、Brave三个组件之间的关系 Brave 是一个 tracer 库，提供的是 tracer 接口Sleuth 采用了 Brave 作为 tracer 库Sleuth 可以不使用 Zipkin

8.6.1.2 Brave 解读

Brave 的两个最基本、也是最核心的概念

trace：以看作是一个逻辑执行过程中的整个链条（可以看作一棵树）span：是 trace 跟踪的基本单位

Brave 中常用的数据结构以及说明

Tracing：工具类，用于生成 Tracer 类实例Tracer：也是工具类，用于生成 SpanSpan：实际记录每个功能块执行信息的类TraceContext：记录 trace 的执行过程中的元数据信息类Propagation：用于在分布式环境或者跨进程条件下的 trace 跟踪时实现 TraceContext 传递的工具类