You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@servicecomb.apache.org by ni...@apache.org on 2019/01/23 03:49:56 UTC
[servicecomb-website] branch master updated: add
customized-tracing-servicecomb of user guide (#163)
This is an automated email from the ASF dual-hosted git repository.
ningjiang pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/servicecomb-website.git
The following commit(s) were added to refs/heads/master by this push:
new 03c1657 add customized-tracing-servicecomb of user guide (#163)
03c1657 is described below
commit 03c16575196d4950b93e65a21261f6f76f1c24ee
Author: bettermanzzy <zh...@huawei.com>
AuthorDate: Wed Jan 23 11:49:52 2019 +0800
add customized-tracing-servicecomb of user guide (#163)
---
2019-01-23-customized-tracing-with-servicecomb.md | 100 ++++++++++++++++++++++
1 file changed, 100 insertions(+)
diff --git a/2019-01-23-customized-tracing-with-servicecomb.md b/2019-01-23-customized-tracing-with-servicecomb.md
new file mode 100644
index 0000000..fee8b56
--- /dev/null
+++ b/2019-01-23-customized-tracing-with-servicecomb.md
@@ -0,0 +1,100 @@
+---
+title: "ServiceComb + Zipkin : 使用篇——自定义追踪"
+lang: cn
+ref: customized-tracing-with-servicecomb
+permalink: /cn/docs/customized-tracing-with-servicecomb/
+excerpt: "本篇将介绍如何使用 ServiceComb 和 Zipkin 实现自定义追踪"
+last_modified_at: 2019-01-23T09:30:30+08:00
+author: Zhou Zhongyuan
+tags: [zipkin,分布式追踪]
+redirect_from:
+ - /theme-setup/
+---
+
+# 自定义追踪功能
+ServiceComb 支持用户在程序中的指定位置处增加追踪数据,可以实现更细力度的追踪。
+## 使用步骤
+添加依赖
+```
+ <dependency>
+ <groupId>org.apache.servicecomb</groupId>
+ <artifactId>tracing-zipkin</artifactId>
+ </dependency>
+```
+在程序入口或者配置处添加 `@EnableZipkinTracing` 注解
+```
+import org.apache.servicecomb.tracing.zipkin.EnableZipkinTracing;
+
+@SpringBootApplication
+@EnableServiceComb
+@EnableZipkinTracing
+public class CalculatorApplication {
+
+ public static void main(String[] args) {
+ SpringApplication.run(CalculatorApplication.class, args);
+ }
+}
+```
+
+在服务程序中的调用方法处添加 `@Span` 注解
+```
+import org.apache.servicecomb.tracing.Span;
+
+@Service
+public class CalculatorServiceImpl implements CalculatorService {
+
+ /**
+ * {@inheritDoc}
+ */
+ @Span
+ @Override
+ public double calculate(double height, double weight) {
+ if (height <= 0 || weight <= 0) {
+ throw new IllegalArgumentException("Arguments must be above 0");
+ }
+ double heightInMeter = height / 100;
+ double bmi = weight / (heightInMeter * heightInMeter);
+ return roundToOnePrecision(bmi);
+ }
+
+ private double roundToOnePrecision(double value) {
+ return new BigDecimal(value).setScale(1, RoundingMode.HALF_UP).doubleValue();
+ }
+}
+```
+使用限制:自定义跟踪仅支持注解请求线程中的方法调用,且带有`@Span`的类必须是spring管理的bean。
+## 快速演示
+下面使用java-chassis/samples/bmi 程序演示,如何使用自定义追踪功能定位应用程序中的问题?
+1. 正常运行bmi程序,结果如下
+ ![正常结果](https://img-blog.csdnimg.cn/20190122101706113.png)
+ ![zipkin追踪正常情况](https://img-blog.csdnimg.cn/20190122101505915.png)
+2. 在bmi程序的calculator服务的calculate方法处,增加一块进程休眠代码(模拟实际工作中调用当前线程处理其他业务的情景),如下
+ ```
+ public double calculate(double height, double weight) {
+
+ try{
+ Thread.currentThread().sleep(5000);
+ } catch (Exception e){
+
+ }
+
+ if (height <= 0 || weight <= 0) {
+ throw new IllegalArgumentException("Arguments must be above 0");
+ }
+ double heightInMeter = height / 100;
+ double bmi = weight / (heightInMeter * heightInMeter);
+ return roundToOnePrecision(bmi);
+ }
+
+ ```
+3. 运行bmi程序,出现如下异常结果。查看zipkin追踪情况和程序报错信息,可以初步确定问题由 bmi/calculator 服务超时未响应导致
+ ![异常结果](https://img-blog.csdnimg.cn/20190121193215681.png)
+ ![zipkin追踪情况](https://img-blog.csdnimg.cn/20190121195130561.png)
+ ![程序报错信息](https://img-blog.csdnimg.cn/20190122093001881.png)
+4. 使用自定义追踪功能,定位问题具体位置
+ 在calculator服务中配置自定义追踪功能,添加注解 `@Span`在calculator服务的方法上 。运行bmi程序,zipkin追踪耗时情况如下。根据span占用的时间戳,可以确定延时问题出现在calculate方法处,点击span,可查看call.path 获取calculate方法的具体位置。
+ ![zipkin追踪异常情况](https://img-blog.csdnimg.cn/20190121200305278.png)
+ ![span信息信息](https://img-blog.csdnimg.cn/20190121200411404.png)
+## 总结
+从上面的示例可以看出,通过配置ServiceComb的自定义追踪功能,可以实现对服务中调用方法、接口的追踪,实现更细力度化的追踪。这对于我们监控服务内部调用、定位服务中的延时问题等非常有帮助。
+