springboot logback的源码分析请参考这篇，讲的很好：https://juejin.im/post/5cd0cdcc6fb9a0323120a3f4

一、日志常规配置

直接贴出logback-spring.xml，基本满足了对日志的常规需求：
1、控制台日志（带颜色显示）
2、全部日志文件（包括全别日志级别的日志，每天一个，自动压缩）
3、错误日志文件（仅抽取ERROR级别日志，每天一个，自动压缩）

说明：
a. 日志存放目录需要在application.yml中配置: logging: path: /data/logs/${spring.application.name}
b. 此配置文件可以自行扩展，比如增加elk输出、调整日志级别、增加日志过滤等。不在本文讨论访问内。

 

<?xml version="1.0" encoding="UTF-8"?>

<configuration scan="true" scanPeriod="60 seconds" debug="false">
    <!-- 参考SpringBoot默认的logback配置，增加了error日志文件 -->
    <!-- org/springframework/boot/logging/logback/base.xml  -->

    <conversionRule conversionWord="clr" converterClass="org.springframework.boot.logging.logback.ColorConverter" />
    <conversionRule conversionWord="wex" converterClass="org.springframework.boot.logging.logback.WhitespaceThrowableProxyConverter" />
    <conversionRule conversionWord="wEx" converterClass="org.springframework.boot.logging.logback.ExtendedWhitespaceThrowableProxyConverter" />

    <property name="LOG_PATH" value="${LOG_PATH:-${LOG_TEMP:-${java.io.tmpdir:-/tmp}}}"/>
    <property name="CONSOLE_LOG_PATTERN" value="${CONSOLE_LOG_PATTERN:-%clr(%d{${LOG_DATEFORMAT_PATTERN:-yyyy-MM-dd HH:mm:ss.SSS}}){faint} %clr(${LOG_LEVEL_PATTERN:-%5p}) %clr(${PID:- }){magenta} %clr(---){faint} %clr([%15.15t]){faint} %clr(%-40.40logger{39}){cyan} %clr(:){faint} %m%n${LOG_EXCEPTION_CONVERSION_WORD:-%wEx}}"/>
    <property name="FILE_LOG_PATTERN" value="${FILE_LOG_PATTERN:-%d{${LOG_DATEFORMAT_PATTERN:-yyyy-MM-dd HH:mm:ss.SSS}} ${LOG_LEVEL_PATTERN:-%5p} ${PID:- } --- [%t] %-40.40logger{39} : %m%n${LOG_EXCEPTION_CONVERSION_WORD:-%wEx}}"/>

    <!-- 控制台日志 -->
    <appender name="CONSOLE" class="ch.qos.logback.core.ConsoleAppender">
        <encoder>
            <pattern>${CONSOLE_LOG_PATTERN}</pattern>
        </encoder>
    </appender>

    <!-- 全量日志 -->
    <appender name="FILE_ALL" class="ch.qos.logback.core.rolling.RollingFileAppender">
        <encoder>
            <pattern>${FILE_LOG_PATTERN}</pattern>
        </encoder>
        <file>${LOG_PATH}/all.log</file>
        <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
            <fileNamePattern>${LOG_PATH}/all.log.%d{yyyy-MM-dd}.%i.gz</fileNamePattern>
            <maxFileSize>${LOG_FILE_MAX_SIZE:-50MB}</maxFileSize>
            <maxHistory>${LOG_FILE_MAX_HISTORY:-0}</maxHistory>
        </rollingPolicy>
    </appender>

    <!-- 错误日志 -->
    <appender name="FILE_ERROR" class="ch.qos.logback.core.rolling.RollingFileAppender">
        <encoder>
            <pattern>${FILE_LOG_PATTERN}</pattern>
        </encoder>
        <file>${LOG_PATH}/err.log</file>
        <rollingPolicy class="ch.qos.logback.core.rolling.SizeAndTimeBasedRollingPolicy">
            <fileNamePattern>${LOG_PATH}/err.log.%d{yyyy-MM-dd}.%i.gz</fileNamePattern>
            <maxFileSize>${LOG_FILE_MAX_SIZE:-50MB}</maxFileSize>
            <maxHistory>${LOG_FILE_MAX_HISTORY:-0}</maxHistory>
        </rollingPolicy>
        <!-- 过滤出ERROR级别的日志 -->
        <filter class="ch.qos.logback.classic.filter.LevelFilter">
            <level>ERROR</level>
            <onMatch>ACCEPT</onMatch>
            <onMismatch>DENY</onMismatch>
        </filter>
    </appender>

    <!-- 日志总开关 -->
    <root level="INFO">
        <appender-ref ref="CONSOLE" />
        <appender-ref ref="FILE_ALL" />
        <appender-ref ref="FILE_ERROR" />
    </root>

    <!-- 日志过滤 -->
    <logger name="org.apache.catalina.startup.DigesterFactory" level="ERROR"/>
    <logger name="org.apache.catalina.util.LifecycleBase" level="ERROR"/>
    <logger name="org.apache.coyote.http11.Http11NioProtocol" level="WARN"/>
    <logger name="org.apache.sshd.common.util.SecurityUtils" level="WARN"/>
    <logger name="org.apache.tomcat.util.net.NioSelectorPool" level="WARN"/>
    <logger name="org.eclipse.jetty.util.component.AbstractLifeCycle" level="ERROR"/>
    <logger name="org.hibernate.validator.internal.util.Version" level="WARN"/>
</configuration>

二、引入sleuth zipkin是怎么输出traceId的？

一直有个疑问，为什么引入sleuth+zipkin之后，日志中会输出traceId和spanId：

 

2019-06-11 16:16:00.117  INFO [test-project,29ce39ac8a54f753,29ce39ac8a54f753,false] 1636 --- [http-nio-12001-exec-2] ...
2019-06-11 16:16:00.332  INFO [test-project,29ce39ac8a54f753,29ce39ac8a54f753,false] 1636 --- [http-nio-12001-exec-2] ...

从上面logback的配置文件FILE_LOG_PATTERN并未发现相关配置，其中必有妖怪！
通过分析引入的sleuth依赖，终于发现在spring-cloud-sleuth-core-xxx.jar找到，请看这个类

 

org.springframework.cloud.sleuth.autoconfig.TraceEnvironmentPostProcessor

    @Override
    public void postProcessEnvironment(ConfigurableEnvironment environment,
            SpringApplication application) {
        Map<String, Object> map = new HashMap<String, Object>();
        // This doesn't work with all logging systems but it's a useful default so you see
        // traces in logs without having to configure it.
        if (Boolean.parseBoolean(environment.getProperty("spring.sleuth.enabled", "true"))) {
            map.put("logging.pattern.level",
                    "%5p [${spring.zipkin.service.name:${spring.application.name:-}},%X{X-B3-TraceId:-},%X{X-B3-SpanId:-},%X{X-Span-Export:-}]");
        }
        addOrReplace(environment.getPropertySources(), map);
    }

这段代码在项目启动阶段，把环境变量logging.pattern.level替换了，增加了spring.application.name、X-B3-TraceId和X-B3-SpanId。实际上就是替换了环境变量LOG_LEVEL_PATTERN，再回到logback的配置文件FILE_LOG_PATTERN：

 

${FILE_LOG_PATTERN:-%d{${LOG_DATEFORMAT_PATTERN:-yyyy-MM-dd HH:mm:ss.SSS}} ${LOG_LEVEL_PATTERN:-%5p} ${PID:- } --- [%t] %-40.40logger{39} : %m%n${LOG_EXCEPTION_CONVERSION_WORD:-%wEx}}

其中${LOG_LEVEL_PATTERN:-%5p}就是引用了这个环境变量。

到此为止，是不是就完了呢？No~~！增加的spring.application.name、X-B3-TraceId和X-B3-SpanId的值从哪里来呢？继续看代码

 

org.springframework.cloud.sleuth.log.Slf4jCurrentTraceContext

这个类中通过MDC ( Mapped Diagnostic Contexts )的方式把traceId、spanId插入到日志内容中。