开发中经常需要根据日志排查问题或跟踪调用流程,很多业务日志并没有考虑排查问题时的便利性,看似都记录了日志,但同一个请求链路的日志无法对应,特别是当日志跨服务时候,或者同一个业务逻辑同一时刻有多条日志,根本无法对应起来,如果日志记可以追踪的话,可以根据全局唯一id搜索得出一条调用链的日志,顺着这个日志链条就可以看出程序的执行全过程,进而有利于排查出问题。
这里我们用两种方式实现:
自己写代码实现:轻量级,灵活,可任意修改,适合小项目;目前已实现支持已spring为基础的springboot,springcloud,dubbo且使用logback日志框架的项目。
使用开源工具:安装费点事,功能强大,适合大项目;支持的开发语言和框架较多。
1. 代码实现
不想一步步的自己实现的话,有已经封装好的jar(建议www.search.maven.org搜最新版本):
<dependency>
<groupId>com.wuyunonline.tracelog</groupId>
<artifactId>tracelog-spring-boot-starter</artifactId>
<version>1.0.0</version>
</dependency>
使用说明和源码地址:https://gitee.com/jwb-wuyun/traclog
下面是自己实现的步骤
1.1 springmvc
先写一个拦截器:
import com.tracelog.common.constant.TraceLogConstant;
import com.tracelog.common.util.TraceIdUtil;
import org.slf4j.MDC;
import org.springframework.util.StringUtils;
import org.springframework.web.servlet.handler.HandlerInterceptorAdapter;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;
public class TraceLogWebMvcInterceptor extends HandlerInterceptorAdapter {
@Override
public boolean preHandle(HttpServletRequest request, HttpServletResponse response, Object handler) {
String method = request.getMethod();
// 放行跨域预请求
if (method.equals("OPTIONS")) {
return true;
}
String traceId = request.getHeader(TraceLogConstant.TRACE_ID);
if (StringUtils.isEmpty(traceId)) {
traceId = TraceIdUtil.uuid_timestamp();
}
MDC.put(TraceLogConstant.TRACE_ID, traceId);
return true;
}
}
再配置拦截器:
import com.tracelog.interceptor.TraceLogWebMvcInterceptor;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.web.servlet.config.annotation.InterceptorRegistry;
import org.springframework.web.servlet.config.annotation.WebMvcConfigurer;
@Configuration
public class TraceLogWebMvcConfig implements WebMvcConfigurer {
@Override
public void addInterceptors(InterceptorRegistry registry) {
registry.addInterceptor(mvcInterceptor()).addPathPatterns("/**");
}
@Bean
public TraceLogWebMvcInterceptor mvcInterceptor() {
return new TraceLogWebMvcInterceptor();
}
}
1.2 springcloud
springcloud一般都是用feign,所以我们需要拦截feign。
先写拦截器:
import com.tracelog.common.constant.TraceLogConstant;
import feign.RequestInterceptor;
import feign.RequestTemplate;
import org.slf4j.MDC;
public class TraceLogFeignInterceptor implements RequestInterceptor {
@Override
public void apply(RequestTemplate requestTemplate) {
requestTemplate.header(TraceLogConstant.TRACE_ID, MDC.get(TraceLogConstant.TRACE_ID));
}
}
再配置拦截器:
import com.tracelog.interceptor.TraceLogFeignInterceptor;
import feign.RequestInterceptor;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
@Configuration
public class TraceLogFeignConfig {
@Bean
public RequestInterceptor requestInterceptor() {
return new TraceLogFeignInterceptor();
}
}
1.3 dubbo
dubbo必须是Apache的dubbo(老版的是阿里的,阿里已将dubbo加入apache)
先写过滤器:
import com.tracelog.common.constant.TraceLogConstant;
import com.tracelog.common.util.TraceIdUtil;
import org.apache.dubbo.common.constants.CommonConstants;
import org.apache.dubbo.common.extension.Activate;
import org.apache.dubbo.rpc.*;
import org.slf4j.MDC;
import org.springframework.util.StringUtils;
@Activate(group = {CommonConstants.PROVIDER, CommonConstants.CONSUMER})
public class TraceLogDubboFilter implements Filter {
@Override
public Result invoke(Invoker<?> invoker, Invocation invocation) throws RpcException {
// 获取dubbo上下文中的traceId
String traceId = invocation.getAttachment(TraceLogConstant.TRACE_ID);
if (StringUtils.isEmpty(traceId)) {
// customer 获取上游来的traceId,并设置到dubbo的上下文,如果没有则生成一个
traceId = MDC.get(TraceLogConstant.TRACE_ID);
if (StringUtils.isEmpty(traceId)) {
traceId = TraceIdUtil.uuid_timestamp();
MDC.put(TraceLogConstant.TRACE_ID, traceId);
}
// provider 设置traceId到日志到上下文
invocation.setAttachment(TraceLogConstant.TRACE_ID, traceId);
} else {
MDC.put(TraceLogConstant.TRACE_ID, traceId);
}
Result result = invoker.invoke(invocation);
return result;
}
}
声明过滤器:
在resource下创建Filter文件,里面内容:traceLogDubboFilter=.TraceLogDubboFilter,是你的实际路径,如下图:
1.4 注解
比如定时任务等利用AOP加切面:@Scheduled,@PostConstruct
import com.tracelog.common.constant.TraceLogConstant;
import com.tracelog.common.util.TraceIdUtil;
import org.aspectj.lang.ProceedingJoinPoint;
import org.aspectj.lang.annotation.Around;
import org.aspectj.lang.annotation.Aspect;
import org.aspectj.lang.annotation.Pointcut;
import org.slf4j.MDC;
import org.springframework.stereotype.Component;
@Aspect
@Component
public class TracelogAspect {
@Pointcut("@annotation(org.springframework.scheduling.annotation.Scheduled) || @annotation(javax.annotation.PostConstruct)")
public void tracelogPointCut() {
}
@Around("tracelogPointCut()")
public Object around(ProceedingJoinPoint point) throws Throwable {
MDC.put(TraceLogConstant.TRACE_ID, TraceIdUtil.uuid_timestamp());
return point.proceed();
}
}
1.5 异步线程获取主线程traceId
针对注解@Async,需要重写线程池,其他线程的处理可以参考这种方式。
先实现修饰类:
import org.slf4j.MDC;
import org.springframework.core.task.TaskDecorator;
import java.util.Map;
public class TraceLogMdcTaskDecorator implements TaskDecorator {
@Override
public Runnable decorate(Runnable runnable) {
Map<String, String> map = MDC.getCopyOfContextMap();
return () -> {
try {
if (map != null) {
MDC.setContextMap(map);
}
runnable.run();
} finally {
MDC.clear();
}
};
}
}
再配置:
import com.tracelog.interceptor.TraceLogMdcTaskDecorator;
import org.springframework.context.annotation.Configuration;
import org.springframework.scheduling.annotation.AsyncConfigurerSupport;
import org.springframework.scheduling.concurrent.ThreadPoolTaskExecutor;
import java.util.concurrent.Executor;
@Configuration
public class TraceLogAsyncConfig extends AsyncConfigurerSupport {
@Override
public Executor getAsyncExecutor() {
ThreadPoolTaskExecutor threadPoolTaskExecutor = new ThreadPoolTaskExecutor();
threadPoolTaskExecutor.setCorePoolSize(50);
threadPoolTaskExecutor.setMaxPoolSize(100);
threadPoolTaskExecutor.setQueueCapacity(1000);
threadPoolTaskExecutor.setThreadNamePrefix("Async-");
threadPoolTaskExecutor.setTaskDecorator(new TraceLogMdcTaskDecorator());
threadPoolTaskExecutor.initialize();
return threadPoolTaskExecutor;
}
}
1.6 扩展支持httpclinet,okhttp
使用Okhttp或HttpClient调用时候如果想传递traceId,必须在自己的Okhttp或HttpClient中加入traclog 拦截器,如下:
Okhttp
实现拦截器:
import com.tracelog.common.constant.TraceLogConstant;
import okhttp3.Interceptor;
import okhttp3.Request;
import okhttp3.Response;
import org.jetbrains.annotations.NotNull;
import org.slf4j.MDC;
import java.io.IOException;
public class TraceLogOkhttpInterceptor implements Interceptor {
@Override
public Response intercept(@NotNull Chain chain) throws IOException {
Request request = chain.request().newBuilder()
.addHeader(TraceLogConstant.TRACE_ID, MDC.get(TraceLogConstant.TRACE_ID))
.build();
return chain.proceed(request);
}
}
使用拦截器
import com.tracelog.interceptor.TraceLogOkhttpInterceptor;
...
OkHttpClient okHttpClient = new OkHttpClient.Builder().addInterceptor(new TraceLogOkhttpInterceptor()).build();
HttpClient
实现拦截器:
import com.tracelog.common.constant.TraceLogConstant;
import org.apache.http.HttpRequest;
import org.apache.http.HttpRequestInterceptor;
import org.apache.http.protocol.HttpContext;
import org.slf4j.MDC;
public class TraceLogHttpClientInterceptor implements HttpRequestInterceptor {
@Override
public void process(HttpRequest httpRequest, HttpContext httpContext){
httpRequest.addHeader(TraceLogConstant.TRACE_ID, MDC.get(TraceLogConstant.TRACE_ID));
}
}
使用拦截器
import com.tracelog.interceptor.TraceLogHttpClientInterceptor;
...
CloseableHttpClient httpClient = HttpClientBuilder.create().addInterceptorFirst(new TraceLogHttpClientInterceptor()).build();
1.7 配置logback.xml
项目中找到logback.xml或logback-spring.xml, 在日志格式中加入[traceId:%X{traceId}] 例如:
启动项目,查看效果,类似下图说明成功。
2. 采用第三方工具
推荐 skywalking。
2.1 配置pom.xml
<!--skywalking traceId 记录到logback日志,请与安装的服务器版本对应-->
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-trace</artifactId>
<version>8.9.0</version>
</dependency>
<dependency>
<groupId>org.apache.skywalking</groupId>
<artifactId>apm-toolkit-logback-1.x</artifactId>
<version>8.9.0</version>
</dependency>
2.2 配置logback.xml
<!-- 日志输出格式 -->
<property name="log.pattern" value="[%tid] %d{yyyy-MM-dd HH:mm:ss.SSS} [%thread] %-5level %logger{36} - [%method,%line] - %msg%n"/>
<!-- 系统日志输出 -->
<appender name="file_info" class="ch.qos.logback.core.rolling.RollingFileAppender">
<file>${log.path}/info.log</file>
<encoder class="ch.qos.logback.core.encoder.LayoutWrappingEncoder">
<layout class="org.apache.skywalking.apm.toolkit.log.logback.v1.x.TraceIdPatternLogbackLayout">
<pattern>${log.pattern}</pattern>
</layout>
<pattern>${log.pattern}</pattern>
</encoder>
</appender>
<!-- 日志收集 skywalking 8.4.0版本开始支持 -->
<appender name="grpcLog" class="org.apache.skywalking.apm.toolkit.log.logback.v1.x.log.GRPCLogClientAppender">
<encoder class="ch.qos.logback.core.encoder.LayoutWrappingEncoder">
<layout class="org.apache.skywalking.apm.toolkit.log.logback.v1.x.mdc.TraceIdMDCPatternLogbackLayout">
<pattern>${log.pattern}</pattern>
</layout>
</encoder>
</appender>
日志收集不是必须的,如果觉得不需要集中收集在一起,那么不需要加GRPCLogClientAppender
注意下面图中的红色标注
2.3 skywalking安装和使用
教程:https://www.jianshu.com/p/b69bc629f476
2.4 效果展示
确保项目中已配置,skywalking已安装,探针也已配置。
log文件和skywalking都可以通过TID(traceId)追踪日志链。
在控制台或log文件中会打印带TID(traceId)的log:
skywalking服务端会收集到带TID(traceId)的log: