一、logback简介

logback是log4j创始人写的,性能比log4j要好,目前主要分为3个模块

  1. logback-core:核心代码模块
  2. logback-classic:log4j的一个改良版本,同时实现了slf4j的接口,这样你如果之后要切换其他日志组件也是一件很容易的事
  3. logback-access:访问模块与Servlet容器集成提供通过Http来访问日志的功能

二、logback.xml配置

<?xml version="1.0" encoding="UTF-8"?>
<!--debug 要不要打印 logback内部日志信息,true则表示要打印。建议开启-->
<!--scan 配置发送改变时,要不要重新加载-->
<configuration debug="true" scan="true" scanPeriod="1 seconds">
    <contextName>logback</contextName>
    <!--定义参数,后面可以通过${app.name}使用-->
    <property name="app.name" value="logback_test"/>
    <!--ConsoleAppender 用于在屏幕上输出日志-->
    <appender name="stdout" class="ch.qos.logback.core.ConsoleAppender">
        <!--定义了一个过滤器,在LEVEL之下的日志输出不会被打印出来-->
        <!--这里定义了DEBUG,也就是控制台不会输出比ERROR级别小的日志-->
        <filter class="ch.qos.logback.classic.filter.ThresholdFilter">
            <level>DEBUG</level>
        </filter>
        <!-- encoder 默认配置为PatternLayoutEncoder -->
        <!--定义控制台输出格式-->
        <encoder>
            <pattern>%d [%thread] %-5level %logger{36} [%file : %line] - %msg%n</pattern>
        </encoder>
    </appender>

    <appender name="file" class="ch.qos.logback.core.rolling.RollingFileAppender">
        <!--定义日志输出的路径-->
        <!--这里的scheduler.manager.server.home 没有在上面的配置中设定,所以会使用java启动时配置的值-->
        <!--比如通过 java -Dscheduler.manager.server.home=/path/to XXXX 配置该属性-->
        <file>${scheduler.manager.server.home}/logs/${app.name}.log</file>
        <!--定义日志滚动的策略TimeBasedRollingPolicy: 最常用的滚动策略,它根据时间来制定滚动策略,
          既负责滚动也负责出发滚动。有以下子节点:<fileNamePattern>:必要节点,包含文件名及“%d”转换符, 
          “%d”可以包含一个java.text.SimpleDateFormat指定的时间格式,如:%d{yyyy-MM}。如果直接使用 %d-->
        <rollingPolicy class="ch.qos.logback.core.rolling.TimeBasedRollingPolicy">
            <!--定义文件滚动时的文件名的格式-->
            <fileNamePattern>${scheduler.manager.server.home}/logs/${app.name}.%d{yyyy-MM-dd.HH}.log.gz
            </fileNamePattern>
            <!--60天的时间周期,日志量最大20GB-->
            <maxHistory>60</maxHistory>
            <!-- 该属性在 1.1.6版本后 才开始支持-->
            <totalSizeCap>20GB</totalSizeCap>
        </rollingPolicy>
        <triggeringPolicy class="ch.qos.logback.core.rolling.SizeBasedTriggeringPolicy">
            <!--每个日志文件最大100MB-->
            <maxFileSize>100MB</maxFileSize>
        </triggeringPolicy>
        <!--定义输出格式-->
        <encoder>
            <pattern>%d [%thread] %-5level %logger{36} [%file : %line] - %msg%n</pattern>
        </encoder>
    </appender>

    <!--root是默认的logger 这里设定输出级别是debug-->
    <root level="info">
        <!--定义了两个appender,日志会通过往这两个appender里面写-->
        <appender-ref ref="stdout"/>
        <appender-ref ref="file"/>
    </root>
    <!--对于类路径以 com.example.logback 开头的Logger,输出级别设置为warn,并且只输出到控制台-->
    <!--这个logger没有指定appender,它会继承root节点中定义的那些appender-->
    <logger name="com.example.logback" level="warn"/>

    <!--通过 LoggerFactory.getLogger("mytest") 可以获取到这个logger-->
    <!--由于这个logger自动继承了root的appender,root中已经有stdout的appender了,自己这边又引入了stdout的appender-->
    <!--如果没有设置 additivity="false" ,就会导致一条日志在控制台输出两次的情况-->
    <!--additivity表示要不要使用rootLogger配置的appender进行输出-->
    <logger name="mytest" level="info" additivity="false">
        <appender-ref ref="stdout"/>
    </logger>
    
    <!--由于设置了 additivity="false" ,所以输出时不会使用rootLogger的appender-->
    <!--但是这个logger本身又没有配置appender,所以使用这个logger输出日志的话就不会输出到任何地方-->
    <logger name="mytest2" level="info" additivity="false"/>
</configuration>

 

三、实现原理

1、获取LoggerFactory

slf4j委托具体实现框架的StaticLoggerBinder来返回一个ILoggerFactory,从而对接到具体实现框架上,我们看下这个类(省略了部分代码)

public class StaticLoggerBinder implements LoggerFactoryBinder {

  private static StaticLoggerBinder SINGLETON = new StaticLoggerBinder();static {
    SINGLETON.init();
  }public static StaticLoggerBinder getSingleton() {
    return SINGLETON;
  }/**
   * Package access for testing purposes.
   */
  void init() {
    try {
      try {
        new ContextInitializer(defaultLoggerContext).autoConfig();
      } catch (JoranException je) {
        Util.report("Failed to auto configure default logger context", je);
      }
      // logback-292
      if(!StatusUtil.contextHasStatusListener(defaultLoggerContext)) {
        StatusPrinter.printInCaseOfErrorsOrWarnings(defaultLoggerContext);
      }
      contextSelectorBinder.init(defaultLoggerContext, KEY);
      initialized = true;
    } catch (Throwable t) {
      // we should never get here
      Util.report("Failed to instantiate [" + LoggerContext.class.getName()
          + "]", t);
    }
  }
  public ILoggerFactory getLoggerFactory() {
    if (!initialized) {
      return defaultLoggerContext;
    }

    if (contextSelectorBinder.getContextSelector() == null) {
      throw new IllegalStateException(
          "contextSelector cannot be null. See also " + NULL_CS_URL);
    }
    return contextSelectorBinder.getContextSelector().getLoggerContext();
  }
}

可以看到

  • 1、通过getSingleton()获取该类的单例
  • 2、static块来保证初始化调用init()方法
  • 3、在init方法中,委托ContextInitializer类对LoggerContext进行初始化。这里如果找到了任一配置文件,就会根据配置文件去初始化LoggerContext,如果没找到,会使用默认配置。
  • 4、然后初始化ContextSelectorStaticBinder,在这个类内部new一个DefaultContextSelector,并把第一步中配置完毕的LoggerContext传给DefaultContextSelector
  • 5、调用getLoggerFactory()方法,直接返回3中配置的LoggerContext,或者委托DefaultContextSelector类返回LoggerContext

这里可以看出所有的配置均保存在LoggerContext这个类中,只要获取到了该类,就能得到log的所有配置,我们的logger就保存在该类的Map<String, Logger> loggerCache中,key为logger的name.

2、获取logger

public final Logger getLogger(final String name) {

        if (name == null) {
            throw new IllegalArgumentException("name argument cannot be null");
        }

        // 如果请求的是ROOT Logger,那么就直接返回root
        if (Logger.ROOT_LOGGER_NAME.equalsIgnoreCase(name)) {
            return root;
        }

        int i = 0;
        Logger logger = root;

        // 请求的Logger是否已经创建过了,如果已经创建过,就直接从loggerCache中返回
        Logger childLogger = (Logger) loggerCache.get(name);
        // if we have the child, then let us return it without wasting time
        if (childLogger != null) {
            return childLogger;
        }

        // if the desired logger does not exist, them create all the loggers
        // in between as well (if they don't already exist)
        String childName;
        while (true) {
            int h = LoggerNameUtil.getSeparatorIndexOf(name, i);
            if (h == -1) {
                childName = name;
            } else {
                childName = name.substring(0, h);
            }
            i = h + 1;
            synchronized (logger) {
                childLogger = logger.getChildByName(childName);
                if (childLogger == null) {
                    //创建Logger实例
                    childLogger = logger.createChildByName(childName);
                    loggerCache.put(childName, childLogger);
                    incSize();
                }
            }
            logger = childLogger;
            if (h == -1) {
                return childLogger;
            }
        }
    }

3、logger.info()记录日志

slf4j定义了Logger接口记录日志的方法是info()、warn()、debug()等,这些方法只是入口,logback是这样实现这些方法的 

public void info(String msg) {  
    filterAndLog_0_Or3Plus(FQCN, null, Level.INFO, msg, null, null);  
  }

 当客户端代码调用Logger.info()时,实际上会进入filterAndLog_0_Or3Plus方法,Logger类中还有很多名字很相似的方法,比如filterAndLog_1、filterAndLog_2。

/** 
   * The next methods are not merged into one because of the time we gain by not 
   * creating a new Object[] with the params. This reduces the cost of not 
   * logging by about 20 nanoseconds. 
   */  
  
  private final void filterAndLog_0_Or3Plus(final String localFQCN,  
      final Marker marker, final Level level, final String msg,  
      final Object[] params, final Throwable t) {  
  
    final FilterReply decision = loggerContext.getTurboFilterChainDecision_0_3OrMore(marker, this, level, msg,  params, t);  
    if (decision == FilterReply.NEUTRAL) {  
      if (effectiveLevelInt > level.levelInt) {  
        return;  
      }  
    } else if (decision == FilterReply.DENY) {  
      return;  
    }  
  
    buildLoggingEventAndAppend(localFQCN, marker, level, msg, params, t);  
  }

该方法首先要请求TurboFilter来判断是否允许记录这次日志信息。TurboFilter是快速筛选的组件,筛选发生在LoggingEvent创建之前,这种设计也是为了提高性能 

如果经过过滤,确定要记录这条日志信息,则进入buildLoggingEventAndAppend方法 

private void buildLoggingEventAndAppend(final String localFQCN,  
      final Marker marker, final Level level, final String msg,  
      final Object[] params, final Throwable t) {  
    LoggingEvent le = new LoggingEvent(localFQCN, this, level, msg, t, params);  
    le.setMarker(marker);  
    callAppenders(le);  
  }

在这个方法里,首先创建了LoggingEvent对象,然后调用callAppenders()方法,要求该Logger关联的所有Appenders来记录日志 

LoggingEvent对象是承载了日志信息的类,最后输出的日志信息,就来源于这个事件对象 

/** 
   * Invoke all the appenders of this logger. 
   *  
   * @param event 
   *          The event to log 
   */  
  public void callAppenders(ILoggingEvent event) {  
    int writes = 0;  
    for (Logger l = this; l != null; l = l.parent) {  
      writes += l.appendLoopOnAppenders(event);  
      if (!l.additive) {  
        break;  
      }  
    }  
    // No appenders in hierarchy  
    if (writes == 0) {  
      loggerContext.noAppenderDefinedWarning(this);  
    }  
  }

经过前面的Filter过滤、日志级别匹配、创建LoggerEvent对象,终于进入了记录日志的方法。该方法会调用此Logger关联的所有Appender,而且还会调用所有父Logger关联的Appender,直到遇到父Logger的additive属性设置为false为止,这也是为什么如果子Logger和父Logger都关联了同样的Appender,则日志信息会重复记录的原因 

private int appendLoopOnAppenders(ILoggingEvent event) {  
    if (aai != null) {  
      return aai.appendLoopOnAppenders(event);  
    } else {  
      return 0;  
    }  
  }

实际上调用的AppenderAttachableImpl的appendLoopOnAppenders()方法 

/** 
   * Call the <code>doAppend</code> method on all attached appenders. 
   */  
  public int appendLoopOnAppenders(E e) {  
    int size = 0;  
    r.lock();  
    try {  
      for (Appender<E> appender : appenderList) {  
        appender.doAppend(e);  
        size++;  
      }  
    } finally {  
      r.unlock();  
    }  
    return size;  
  }

到这里,为了记录一条日志信息,长长的调用链终于告一段落了,通过调用Appender的doAppend(LoggingEvent e)方法,委托Appender来最终记录日志

UnsynchronizedAppenderBase里面的doAppend()方法,它主要是记录了Status状态,然后检查Appender上的Filter是否满足过滤条件,最后再调用实现子类的appender()方法。很眼熟是吗,这里用到了一个设计模式——模板方法  

public void doAppend(E eventObject) {  
    // WARNING: The guard check MUST be the first statement in the  
    // doAppend() method.  
        
    // prevent re-entry.  
    if (Boolean.TRUE.equals(guard.get())) {  
      return;  
    }  
  
    try {  
      guard.set(Boolean.TRUE);  
  
      if (!this.started) {  
        if (statusRepeatCount++ < ALLOWED_REPEATS) {  
          addStatus(new WarnStatus(  
              "Attempted to append to non started appender [" + name + "].",  
              this));  
        }  
        return;  
      }  
  
      if (getFilterChainDecision(eventObject) == FilterReply.DENY) {  
        return;  
      }  
  
      // ok, we now invoke derived class' implementation of append  
      this.append(eventObject);  
  
    } catch (Exception e) {  
      if (exceptionCount++ < ALLOWED_REPEATS) {  
        addError("Appender [" + name + "] failed to append.", e);  
      }  
    } finally {  
      guard.set(Boolean.FALSE);  
    }  
  }  
  
  abstract protected void append(E eventObject);

上面的代码非常简单,就不用说了,我们就直接看看实现类的append()方法是怎么实现的,这里我们选择OutputStreamAppender实现类 

@Override  
  protected void append(E eventObject) {  
    if (!isStarted()) {  
      return;  
    }  
  
    subAppend(eventObject);  
  }

首先检查一下这个Appender是否已经启动,如果没启动就直接返回,如果已经启动,则又进入一个subAppend()方法 

RollingFileAppender覆盖了subAppend()方法,实现了翻滚策略

@Override
    protected void subAppend(E event) {
        // The roll-over check must precede actual writing. This is the
        // only correct behavior for time driven triggers.

        // We need to synchronize on triggeringPolicy so that only one rollover
        // occurs at a time
        synchronized (triggeringPolicy) {
            if (triggeringPolicy.isTriggeringEvent(currentlyActiveFile, event)) {
                rollover();
            }
        }

        super.subAppend(event);
    }

进入super.subAppend(event);

protected void subAppend(E event) {
        if (!isStarted()) {
            return;
        }
        try {
            // this step avoids LBCLASSIC-139
            if (event instanceof DeferredProcessingAware) {
                ((DeferredProcessingAware) event).prepareForDeferredProcessing();
            }
            // the synchronization prevents the OutputStream from being closed while we
            // are writing. It also prevents multiple threads from entering the same
            // converter. Converters assume that they are in a synchronized block.
            // lock.lock();

            byte[] byteArray = this.encoder.encode(event);
            writeBytes(byteArray);

        } catch (IOException ioe) {
            // as soon as an exception occurs, move to non-started state
            // and add a single ErrorStatus to the SM.
            this.started = false;
            addStatus(new ErrorStatus("IO failure in appender", this, ioe));
        }
    }

通过this.encoder.encode(event)方法格式化需要记录的日志,然后通过writeBytes(byteArray)写入日志。

 

四、通过代码动态生成logger对象

public class LoggerHolder {

    public static Logger getLogger(String name) {
        LoggerContext loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory();
        //如果未创建该logger
        if (loggerContext.exists(name) == null) {
            return buildLogger(name);
        }
        //如果已经创建,则返回
        return loggerContext.getLogger(name);
    }

    private static Logger buildLogger(String name) {

        LoggerContext loggerContext = (LoggerContext) LoggerFactory.getILoggerFactory();
        Logger logger = loggerContext.getLogger(name);
        //配置rollingFileAppender
        RollingFileAppender rollingFileAppender = new RollingFileAppender();
        rollingFileAppender.setName(name);
        //配置rollingPolicy
        TimeBasedRollingPolicy rollingPolicy = new TimeBasedRollingPolicy();
        rollingPolicy.setFileNamePattern("/data/pjf/" + name + "/" + name + ".%d{yyyyMMdd}.log");
        rollingFileAppender.setRollingPolicy(rollingPolicy);
        //配置encoder
        PatternLayoutEncoder encoder = new PatternLayoutEncoder();
        encoder.setCharset(UTF_8);
        encoder.setPattern("%msg%n");
        rollingFileAppender.setEncoder(encoder);
        //配置logger
        logger.addAppender(rollingFileAppender);
        logger.setAdditive(false);
        logger.setLevel(Level.INFO);
        return logger;
    }
}

 

具体参考:写的很详细,

读logback源码系列文章(一)——对接slf4j
读logback源码系列文章(二)——提供ILoggerFactory
读logback源码系列文章(三)——创建Logger
读logback源码系列文章(四)——记录日志
读logback源码系列文章(五)——Appender
读logback源码系列文章(六)——ContextInitializer
读logback源码系列文章(七)——配置的实际工作类Action
读logback源码系列文章(八)——记录日志的实际工作类Encoder