Sleuth 使用

sleuth是spring cloud中日志链（调用链解决方案）,自动添加traceId，spanId。

sleuth包也可单独引入，和其它藕合度小。sleuth + zipkin 实现APM管理。

Sleuth 引入文档

从其它文章web压测结果获悉：sleuth在webmvc下没多大损失，在webflux下损失较大。需注意。

参见：Spring Cloud Sleuth 性能测试_Apple_zhl的博客-CSDN博客

MDC写入信息是在Slf4jScopeDecorator中。

MDC会放入以下内容：

final String previousTraceId = MDC.get("traceId");
final String previousParentId = MDC.get("parentId");
final String previousSpanId = MDC.get("spanId");
final String spanExportable = MDC.get("spanExportable");
final String legacyPreviousTraceId = MDC.get(LEGACY_TRACE_ID_NAME);
final String legacyPreviousParentId = MDC.get(LEGACY_PARENT_ID_NAME);
final String legacyPreviousSpanId = MDC.get(LEGACY_SPAN_ID_NAME);
final String legacySpanExportable = MDC.get(LEGACY_EXPORTABLE_NAME);
private static final String LEGACY_EXPORTABLE_NAME = "X-Span-Export";
private static final String LEGACY_PARENT_ID_NAME = "X-B3-ParentSpanId";
private static final String LEGACY_TRACE_ID_NAME = "X-B3-TraceId";
private static final String LEGACY_SPAN_ID_NAME = "X-B3-SpanId";

Web访问：

sleuth通过在web的请求增加TraceFilter来开启一个新的spanId 和 traceId，在离开时调用close方法关闭之前的trace.

参见类：

TraceWebServletAutoConfiguration -> TracingFilter

将TracingFilter增加到Web的Filter中，同时设置Filter的Order级别为Integer.MIN + 5.也即基本为最前。

因 TracingFilter是建立在Server级别的， 因此加载顺序会优先于Spring程序在Interceptor上设置的拦截器。

Scheduler:

sleuth使用AOP切面拦截所有的scheduler，使用TraceSchedulingAspect类控制生成新的traceId

WebClient数据传递：

TraceRestTemplateCustomizer -> TracingClientHttpRequestInterceptor

因此，只要使用RestTemplate.Builder生成的信息，都会默认会执行RestTemplateCustomizer。而自己new RestTemplate出来的对象则没有此功能。

以上都会自动添加。

pom.xml 引入：

<dependency>
<groupId>org.springframework.cloud</groupId>
<artifactId>spring-cloud-sleuth-core</artifactId>
<version>2.1.0.RELEASE</version>
</dependency>

线程和线程池使用：

线程-依托TraceRunnable 或 TraceCallable：

public class ThreadTestController {

@Autowired
private Tracing tracing;

@Autowired
private SpanNamer spanNamer;

@RequestMapping("/thread")
public String starterThread() throws InterruptedException {
log.info("trace parent...");
Runnable runnable = new Runnable() {
@Override
public void run() {
log.info("thread test");
}
};
TraceRunnable traceRunnable = new TraceRunnable(tracing, spanNamer, runnable, "myRunnable");
//TraceCallable
Thread thread = new Thread(traceRunnable);
thread.start();
return"OK";
}
}

线程池依托：

内部调用

@Autowired
private BeanFactory beanFactory;

ThreadPoolExecutor myThreadPool = new ThreadPoolExecutor(2, 2,0L,TimeUnit.MILLISECONDS,new LinkedBlockingQueue<Runnable>());

@RequestMapping("/executor2")
public String executor2() throws InterruptedException {
TraceableExecutorService executorService = new TraceableExecutorService(beanFactory, myThreadPool, "1234");
log.info("executor2 parent...");
//TraceCallable

//TraceCallable
for(int i=0;i< 30;i++){
executorService.submit(new SpanRunnable());
}
return"OK";
}
}

bean定义：

@Configuration
@Slf4j
public class BeanConfiguration {

/**
* 切换链接池
* @param beanFactory
* @return
*/
@Bean("myTestExecutor")
public ExecutorService executorService(@Autowired BeanFactory beanFactory){
ThreadPoolExecutor poolExecutor = new ThreadPoolExecutor(5, 5, 100, TimeUnit.SECONDS, new LinkedBlockingQueue<Runnable>());
log.info("poolExecutor:--------------{}",poolExecutor);
TraceableExecutorService executorService = new TraceableExecutorService(beanFactory, poolExecutor, "executePool");
return executorService;
    }
}

@Autowired
@Qualifier("myTestExecutor")
private ExecutorService executorService;

/**
*
* @return
* @throws InterruptedException
*/
@RequestMapping("/executor3")
public String executor3() throws InterruptedException {
log.info("executor3 parent...");
log.info("taskExecutor:{}", taskExecutor);
//TraceCallable

//TraceCallable
for(int i=0;i< 30;i++){
executorService.submit(new SpanRunnable());
}
return"OK";
}