放入|时会_性能优化之异步日志

篇首语：本文由编程笔记#小编为大家整理，主要介绍了性能优化之异步日志相关的知识，希望对你有一定的参考价值。

          你听说过打印日志能把系统拖垮的情况吗。

          所以项目中一般打印日志会使用异步AsyncAppender打印日志。

          那么使用了AsyncAppender&＃xff0c;会不会性能就好了&＃xff0c;就不会阻塞业务流程了&＃xff0c;会不会丢失日志呢&＃xff0c;我们来看一下logback的实现。

          先看下官方文档的介绍&＃xff0c;AsyncAppender 会把处理的事件缓存到一个阻塞队列&＃xff0c;默认情况下达到队列容量的80%的时候&＃xff0c;会丢弃TRACE, DEBUG and INFO级别的事件&＃xff0c;

根据默认配置neverBlock&＃61;false&＃xff0c;队列也会发生阻塞&＃xff0c;所以设置true&＃xff0c;虽然完全是非阻塞但会丢失日志&＃xff0c;特别是error级别的日志。

       我们看下源码如何实现的。

public class AsyncAppender extends AsyncAppenderBase
boolean includeCallerData &＃61; false;
/**
* Events of level TRACE, DEBUG and INFO are deemed to be discardable.
* &＃64;param event
* &＃64;return true if the event is of level TRACE, DEBUG or INFO false otherwise.
*/
protected boolean isDiscardable(ILoggingEvent event)
Level level &＃61; event.getLevel();
return level.toInt() <&＃61; Level.INFO_INT;

protected void preprocess(ILoggingEvent eventObject)
eventObject.prepareForDeferredProcessing();
if (includeCallerData)
eventObject.getCallerData();

public boolean isIncludeCallerData()
return includeCallerData;

public void setIncludeCallerData(boolean includeCallerData)
this.includeCallerData &＃61; includeCallerData;



根据实现&＃xff1a;

protected boolean isDiscardable(ILoggingEvent event)
Level level &＃61; event.getLevel();
return level.toInt() <&＃61; Level.INFO_INT;


public static final int OFF_INT &＃61; Integer.MAX_VALUE;
public static final int ERROR_INT &＃61; 40000;
public static final int WARN_INT &＃61; 30000;
public static final int INFO_INT &＃61; 20000;
public static final int DEBUG_INT &＃61; 10000;
public static final int TRACE_INT &＃61; 5000;
public static final int ALL_INT &＃61; Integer.MIN_VALUE;

TRACE, DEBUG and INFO级别的事件&＃xff0c;在容量默认到达80%的时候会丢弃。

事件放入阻塞队列的时候即调用put方法时会判断容量及是否可以丢弃条件的判断&＃xff0c;如下&＃xff1a;

&＃64;Override
protected void append(E eventObject)
if (isQueueBelowDiscardingThreshold() && isDiscardable(eventObject))
return;

preprocess(eventObject);
put(eventObject);

private boolean isQueueBelowDiscardingThreshold()
return (blockingQueue.remainingCapacity()
private void put(E eventObject)
if (neverBlock)
blockingQueue.offer(eventObject);
else
putUninterruptibly(eventObject);


private void putUninterruptibly(E eventObject)
boolean interrupted &＃61; false;
try
while (true)
try
blockingQueue.put(eventObject);
break;
catch (InterruptedException e)
interrupted &＃61; true;


finally
if (interrupted)
Thread.currentThread().interrupt();




当事件真正放入队列&＃xff0c;是调用阻塞方法put&＃xff0c;还是非阻塞方法offer&＃xff0c;是由neverBlock配置决定的&＃xff0c;默认false&＃xff0c;会阻塞&＃xff0c;所以为了性能&＃xff0c;我们可以设置为true&＃xff0c;也会丢弃日志。

真正处理队列事件的线程是默无闻的后台&＃xff08;Daemon&＃xff09;消费线程&＃xff0c;

Worker worker &＃61; new Worker();

worker.setDaemon(true);
worker.setName("AsyncAppender-Worker-" &＃43; getName());
// make sure this instance is marked as "started" before staring the worker Thread
super.start();
worker.start()

 事件的最终归宿AppenderAttachableImpl aa。

class Worker extends Thread
public void run()
AsyncAppenderBase parent &＃61; AsyncAppenderBase.this;
AppenderAttachableImpl aai &＃61; parent.aai;
// loop while the parent is started
while (parent.isStarted())
try
E e &＃61; parent.blockingQueue.take();
aai.appendLoopOnAppenders(e);
catch (InterruptedException ie)
break;


addInfo("Worker thread will flush remaining events before exiting. ");
for (E e : parent.blockingQueue)
aai.appendLoopOnAppenders(e);
parent.blockingQueue.remove(e);

aai.detachAndStopAllAppenders();



     异步打印日志是否非阻塞&＃xff0c;是否会丢弃日志&＃xff0c;是完全可以配置的&＃xff0c;如果可以丢弃&＃xff0c;依据日志的报警系统就不太准了&＃xff0c;只能靠打点系统在系统内部aop的方式或其他方式打点上报了。