源将事件写到一个多或者多个通道中。
接收器只从一个通道接收事件。
代理可能会有多个源、通道与接收器。
flume-ng-channels
里面包含了filechannel，jdbcchannel，kafkachannel,memorychannel通道的实现。
flume-ng-clients
实现了log4j相关的几个Appender，使得log4j的日志输出可以直接发送给flume-agent；其中有一个LoadBalancingLog4jAppender的实现，提供了多个flume-agent的load balance和ha功能，采用flume作为日志收集的可以考虑将这个appender引入内部的log4j中。
flume-ng-configuration
这个主要就是Flume配置信息相关的类，包括载入flume-config.properties配置文件并解析。其中包括了Source的配置，Sink的配置，Channel的配置，在阅读源码前推荐先梳理这部分关系再看其他部分的。
flume-ng-core
flume整个核心框架，包括了各个模块的接口以及逻辑关系实现。其中instrumentation是flume内部实现的一套metric机制，metric的变化和维护，其核心也就是在MonitoredCounterGroup中通过一个Map来实现metric的计量。ng-core下几乎大部分代码任然几种在channel、sink、source几个子目录下，其他目录基本完成一个util和辅助的功能。
flume-ng-node
实现启动flume的一些基本类，包括main函数的入口（Application.java中）。在理解configuration之后，从application的main函数入手，可以较快的了解整个flume的代码。

flume-ng启动文件介绍

################################
# constants
################################

#设置常量值，主要是针对不同的参数执行相应的类，以启动Flume环境
FLUME_AGENT_CLASS="org.apache.flume.node.Application"
FLUME_AVRO_CLIENT_CLASS="org.apache.flume.client.avro.AvroCLIClient"
FLUME_VERSION_CLASS="org.apache.flume.tools.VersionInfo"
FLUME_TOOLS_CLASS="org.apache.flume.tools.FlumeToolsMain"

#真正启动Flume环境的方法
run_flume() {
  local FLUME_APPLICATION_CLASS

  if [ "$#" -gt 0 ]; then
    FLUME_APPLICATION_CLASS=$1
    shift
  else
    error "Must specify flume application class" 1
  fi

  if [ ${CLEAN_FLAG} -ne 0 ]; then
    set -x
  fi

  #执行这一行命令，执行相应的启动类，比如org.apache.flume.node.Application
  $EXEC $JAVA_HOME/bin/java $JAVA_OPTS $FLUME_JAVA_OPTS "${arr_java_props[@]}" -cp "$FLUME_CLASSPATH" \
      -Djava.library.path=$FLUME_JAVA_LIBRARY_PATH "$FLUME_APPLICATION_CLASS" $*
}


################################
# main
################################

# set default params
# 在启动的过程中使用到的参数
FLUME_CLASSPATH=""
FLUME_JAVA_LIBRARY_PATH=""
#默认占用堆空间大小，这一块都可以根据JVM进行重新设置
JAVA_OPTS="-Xmx20m"
LD_LIBRARY_PATH=""

opt_conf=""
opt_classpath=""
opt_plugins_dirs=""
arr_java_props=()
arr_java_props_ct=0
opt_dryrun=""

# 根据不同的参数，执行不同的启动类，每个常量所对应的类路径在代码前面有过介绍。
if [ -n "$opt_agent" ] ; then
  run_flume $FLUME_AGENT_CLASS $args
elif [ -n "$opt_avro_client" ] ; then
  run_flume $FLUME_AVRO_CLIENT_CLASS $args
elif [ -n "${opt_version}" ] ; then
  run_flume $FLUME_VERSION_CLASS $args
elif [ -n "${opt_tool}" ] ; then
  run_flume $FLUME_TOOLS_CLASS $args
else
  error "This message should never appear" 1
fi

主要是run-flume语句以及指定的启动类
从bin/flume-ng这个shell脚本可以看到Flume的起始于org.apache.flume.node.Application类，这是flume的main函数所在。main方法首先会先解析shell命令，如果指定的配置文件不存在就抛出异常。
main方法首先校验shell命令行的代码，解析
./bin/flume-ng agent -n agent -c conf -f conf/hw.conf -Dflume.root.logger=INFO,console
语句，然后加载配置文件。
根据命令中含有"no-reload-conf"参数，决定采用那种加载配置文件方式：
一、没有此参数，会动态加载配置文件，默认每30秒加载一次配置文件，因此可以动态修改配置文件；
二、有此参数，则只在启动时加载一次配置文件。实现动态加载功能采用了发布订阅模式，使用guava中的EventBus实现。
三、PropertiesFileConfigurationProvider这个类是配置文件加载类，PollingPropertiesFileConfigurationProvider类中，它实现了LifecycleAware接口，而这个接口是掌管整个Flume生命周期的一个核心接口，LifecycleSupervisor实现了这个接口，通过上面代码中application.start方法触发LifecyleAware的start方法，LifecycleAware接口定义了start和stop方法。
时序调用图：

图片.png

source - channelProcessor - channel - sink -sinkgroup

Source  {
    ChannelProcessor  {
             Channel  ch1
             Channel  ch2
             …
    }
} 
Sink  {
   Channel  ch; 
} 
SinkGroup {
   Channel ch；
   Sink s1；
   Sink s2；
   …
}

1、Source组件
Source是数据源的总称，设定好源后，数据将源源不断的被抓取或者被推送。常见的数据源有：ExecSource，KafkaSource，HttpSource，NetcatSource，JmsSource，AvroSource等等。所有的数据源统一实现一个接口类如下：

@InterfaceAudience.Public
@InterfaceStability.Stable
public interface Source extends LifecycleAware, NamedComponent {

  /**
   * Specifies which channel processor will handle this source's events.
   *
   * @param channelProcessor
   */
  public void setChannelProcessor(ChannelProcessor channelProcessor);

  /**
   * Returns the channel processor that will handle this source's events.
   */
  public ChannelProcessor getChannelProcessor();

}

Source提供了两种机制： PollableSource（轮询拉取）和EventDrivenSource（事件驱动）：

上图展示的Source继承关系类图。通过类图我们可以看到NetcatSource，ExecSource和HttpSource属于事件驱动模型。KafkaSource，SequenceGeneratorSource和JmsSource属于轮询拉取模型。Source接口继承了LifecycleAware接口，它的的所有逻辑的实现在接口的start和stop方法中进行。下图是类关系方法图：

Source接口定义的是最终的实现过程，如抓取日志，这个抓取的过程和实际操作就是在对应的Source实现中，比如：ExecSource。那么这些Source实现由谁来驱动的呢？现在我们将介绍SourceRunner类。看一下类继承结构图：

我们看一下PollableSourceRunner和EventDrivenSourceRunner的具体实现：

  @Override //PollableSourceRunner：
  public void start() {
    PollableSource source = (PollableSource) getSource();
    ChannelProcessor cp = source.getChannelProcessor();
    cp.initialize();
    source.start();

    runner = new PollingRunner();

    runner.source = source;
    runner.counterGroup = counterGroup;
    runner.shouldStop = shouldStop;

    runnerThread = new Thread(runner);
    runnerThread.setName(getClass().getSimpleName() + "-" + 
        source.getClass().getSimpleName() + "-" + source.getName());
    runnerThread.start();

    lifecycleState = LifecycleState.START;
  }

  @Override //EventDrivenSourceRunner
  public void start() {
    Source source = getSource();
    ChannelProcessor cp = source.getChannelProcessor();
    cp.initialize();
    source.start();
    lifecycleState = LifecycleState.START;
  }

注：其实所有的Source实现类内部都维护着线程，执行source.start()其实就是启动了相应的线程。
2、Channel组件
Channel用于连接Source和Sink，Source将日志信息发送到Channel，Sink从Channel消费日志信息；Channel是中转日志信息的一个临时存储，保存有Source组件传递过来的日志信息。先看代码如下：AbstractConfigurationProvider的loadSources()

          ChannelSelectorConfiguration selectorConfig =
              config.getSelectorConfiguration();

          ChannelSelector selector = ChannelSelectorFactory.create(
              sourceChannels, selectorConfig);

          ChannelProcessor channelProcessor = new ChannelProcessor(selector);
          Configurables.configure(channelProcessor, config);

          source.setChannelProcessor(channelProcessor);

ChannelSelectorFactory.create方法实现如下：

  public static ChannelSelector create(List<Channel> channels,
      ChannelSelectorConfiguration conf) {
    String type = ChannelSelectorType.REPLICATING.toString();
    if (conf != null) {
      type = conf.getType();
    }
    ChannelSelector selector = getSelectorForType(type);
    selector.setChannels(channels);
    Configurables.configure(selector, conf);
    return selector;
  }

其中我们看一下ChannelSelectorType这个枚举类，包括了几种类型：

public enum ChannelSelectorType {

  /**
   * Place holder for custom channel selectors not part of this enumeration.
   */
  OTHER(null),

  /**
   * Replicating channel selector.
   */
  REPLICATING("org.apache.flume.channel.ReplicatingChannelSelector"),

  /**
   * Multiplexing channel selector.
   */
  MULTIPLEXING("org.apache.flume.channel.MultiplexingChannelSelector");

ChannelSelector的类结构图如下所示：

注：RelicatingChannelSelector和MultiplexingChannelSelector是二个通道选择器，第一个是复用型通道选择器，也就是的默认的方式，会把接收到的消息发送给其他每个channel。第二个是多路通道选择器，这个会根据消息header中的参数进行通道选择。
Channel是什么？：

@InterfaceAudience.Public
@InterfaceStability.Stable
public interface Channel extends LifecycleAware, NamedComponent {

  /**
   * <p>Puts the given event into the channel.</p>
   * <p><strong>Note</strong>: This method must be invoked within an active
   * {@link Transaction} boundary. Failure to do so can lead to unpredictable
   * results.</p>
   * @param event the event to transport.
   * @throws ChannelException in case this operation fails.
   * @see org.apache.flume.Transaction#begin()
   */
  public void put(Event event) throws ChannelException;

  /**
   * <p>Returns the next event from the channel if available. If the channel
   * does not have any events available, this method must return {@code null}.
   * </p>
   * <p><strong>Note</strong>: This method must be invoked within an active
   * {@link Transaction} boundary. Failure to do so can lead to unpredictable
   * results.</p>
   * @return the next available event or {@code null} if no events are
   * available.
   * @throws ChannelException in case this operation fails.
   * @see org.apache.flume.Transaction#begin()
   */
  public Event take() throws ChannelException;

  /**
   * @return the transaction instance associated with this channel.
   */
  public Transaction getTransaction();
}

注：put方法是用来发送消息，take方法是获取消息，transaction是用于事务操作。类结构图如下：

3、Sink组件
Sink负责取出Channel中的消息数据，进行相应的存储文件系统，数据库，或者提交到远程服务器。Sink在设置存储数据时，可以向文件系统中，数据库中，hadoop中储数据，在日志数据较少时，可以将数据存储在文件系中，并且设定一定的时间间隔保存数据。在日志数据较多时，可以将相应的日志数据存储到Hadoop中，便于日后进行相应的数据分析。
Sink接口类内容如下：

@InterfaceAudience.Public
@InterfaceStability.Stable
public interface Sink extends LifecycleAware, NamedComponent {
  /**
   * <p>Sets the channel the sink will consume from</p>
   * @param channel The channel to be polled
   */
  public void setChannel(Channel channel);

  /**
   * @return the channel associated with this sink
   */
  public Channel getChannel();

  /**
   * <p>Requests the sink to attempt to consume data from attached channel</p>
   * <p><strong>Note</strong>: This method should be consuming from the channel
   * within the bounds of a Transaction. On successful delivery, the transaction
   * should be committed, and on failure it should be rolled back.
   * @return READY if 1 or more Events were successfully delivered, BACKOFF if
   * no data could be retrieved from the channel feeding this sink
   * @throws EventDeliveryException In case of any kind of failure to
   * deliver data to the next hop destination.
   */
  public Status process() throws EventDeliveryException;

  public static enum Status {
    READY, BACKOFF
  }
}

Sink是通过如下代码进行的创建：
Sink sink = sinkFactory.create(comp.getComponentName(), comp.getType());

DefaultSinkFactory.create方法如下：

  @Override
  public Sink create(String name, String type) throws FlumeException {
    Preconditions.checkNotNull(name, "name");
    Preconditions.checkNotNull(type, "type");
    logger.info("Creating instance of sink: {}, type: {}", name, type);
    Class<? extends Sink> sinkClass = getClass(type);
    try {
      Sink sink = sinkClass.newInstance();
      sink.setName(name);
      return sink;
    } catch (Exception ex) {
      throw new FlumeException("Unable to create sink: " + name
          + ", type: " + type + ", class: " + sinkClass.getName(), ex);
    }
  }

**注：Sink是通过SinkFactory工厂来创建，提供了DefaultSinkFactory默认工厂，程序会查找org.apache.flume.conf.sink.SinkType这个枚举类找到相应的Sink处理类，比如：org.apache.flume.sink.LoggerSink，如果没找到对应的处理类，直接通过Class.forName(className)进行直接查找实例化实现类。 **
Sink的类结构图如下：

与ChannelProcessor处理类对应的是SinkProcessor，由SinkProcessorFactory工厂类负责创建，SinkProcessor的类型由一个枚举类提供，看下面代码：

public enum SinkProcessorType {
  /**
   * Place holder for custom sinks not part of this enumeration.
   */
  OTHER(null),

  /**
   * Failover processor
   *
   * @see org.apache.flume.sink.FailoverSinkProcessor
   */
  FAILOVER("org.apache.flume.sink.FailoverSinkProcessor"),

  /**
   * Standard processor
   *
   * @see org.apache.flume.sink.DefaultSinkProcessor
   */
  DEFAULT("org.apache.flume.sink.DefaultSinkProcessor"),


  /**
   * Load balancing processor
   *
   * @see org.apache.flume.sink.LoadBalancingSinkProcessor
   */
  LOAD_BALANCE("org.apache.flume.sink.LoadBalancingSinkProcessor");

SinkProcessor的类结构图如下：

说明：1、FailoverSinkProcessor是故障转移处理器，当sink从通道拿数据信息时出错进行的相关处理，代码如下：

 @Override
  public Status process() throws EventDeliveryException {
    // Retry any failed sinks that have gone through their "cooldown" period
    Long now = System.currentTimeMillis();
    while (!failedSinks.isEmpty() && failedSinks.peek().getRefresh() < now) {
      FailedSink cur = failedSinks.poll();
      Status s;
      try {
        s = cur.getSink().process();
        if (s  == Status.READY) {
          liveSinks.put(cur.getPriority(), cur.getSink());
          activeSink = liveSinks.get(liveSinks.lastKey());
          logger.debug("Sink {} was recovered from the fail list",
                  cur.getSink().getName());
        } else {
          // if it's a backoff it needn't be penalized.
          failedSinks.add(cur);
        }
        return s;
      } catch (Exception e) {
        cur.incFails();
        failedSinks.add(cur);
      }
    }

    Status ret = null;
    while (activeSink != null) {
      try {
        ret = activeSink.process();
        return ret;
      } catch (Exception e) {
        logger.warn("Sink {} failed and has been sent to failover list",
                activeSink.getName(), e);
        activeSink = moveActiveToDeadAndGetNext();
      }
    }

    throw new EventDeliveryException("All sinks failed to process, " +
        "nothing left to failover to");
  }

2、LoadBalancingSinkProcessor是负载Sink处理器首先我们和ChannelProcessor一样，我们也要重点说明一下SinkSelector这个选择器。先看一下LoadBalancingSinkProcessor extends AbstractSinkProcessor的configure方法的部分代码：

  public interface SinkSelector extends Configurable, LifecycleAware {
    void setSinks(List<Sink> sinks);
    Iterator<Sink> createSinkIterator();
    void informSinkFailed(Sink failedSink);
  }

    if (selectorTypeName.equalsIgnoreCase(SELECTOR_NAME_ROUND_ROBIN)) {
      selector = new RoundRobinSinkSelector(shouldBackOff);
    } else if (selectorTypeName.equalsIgnoreCase(SELECTOR_NAME_RANDOM)) {
      selector = new RandomOrderSinkSelector(shouldBackOff);
    } else {
      try {
        @SuppressWarnings("unchecked")
        Class<? extends SinkSelector> klass = (Class<? extends SinkSelector>)
            Class.forName(selectorTypeName);
        selector = klass.newInstance();
      } catch (Exception ex) {
        throw new FlumeException("Unable to instantiate sink selector: "
            + selectorTypeName, ex);
      }
    }

结合上面的代码，再看类结构图如下：

注：RoundRobinSinkSelector是轮询选择器，RandomOrderSinkSelector是随机分配选择器。
以KafkaSink为例看一下Sink里面的具体实现：

public Status process() throws EventDeliveryException {
    Status result = Status.READY;
    Channel channel = getChannel();
    Transaction transaction = null;
    Event event = null;
    String eventTopic = null;
    String eventKey = null;

    try {
      long processedEvents = 0;

      transaction = channel.getTransaction();
      transaction.begin();

      messageList.clear();
      for (; processedEvents < batchSize; processedEvents += 1) {
        event = channel.take();

        if (event == null) {
          // no events available in channel
          break;
        }

        byte[] eventBody = event.getBody();
        Map<String, String> headers = event.getHeaders();

        if ((eventTopic = headers.get(TOPIC_HDR)) == null) {
          eventTopic = topic;
        }

        eventKey = headers.get(KEY_HDR);

        if (logger.isDebugEnabled()) {
          logger.debug("{Event} " + eventTopic + " : " + eventKey + " : "
            + new String(eventBody, "UTF-8"));
          logger.debug("event #{}", processedEvents);
        }

        // create a message and add to buffer
        KeyedMessage<String, byte[]> data = new KeyedMessage<String, byte[]>
          (eventTopic, eventKey, eventBody);
        messageList.add(data);

      }

      // publish batch and commit.
      if (processedEvents > 0) {
        long startTime = System.nanoTime();
        producer.send(messageList);
        long endTime = System.nanoTime();
        counter.addToKafkaEventSendTimer((endTime-startTime)/(1000*1000));
        counter.addToEventDrainSuccessCount(Long.valueOf(messageList.size()));
      }

      transaction.commit();

    } catch (Exception ex) {
      String errorMsg = "Failed to publish events";
      logger.error("Failed to publish events", ex);
      result = Status.BACKOFF;
      if (transaction != null) {
        try {
          transaction.rollback();
          counter.incrementRollbackCount();
        } catch (Exception e) {
          logger.error("Transaction rollback failed", e);
          throw Throwables.propagate(e);
        }
      }
      throw new EventDeliveryException(errorMsg, ex);
    } finally {
      if (transaction != null) {
        transaction.close();
      }
    }

    return result;
  }

注：方法从channel中不断的获取数据，然后通过Kafka的producer生产者将消息发送到Kafka里面

主要参考：http://www.jianshu.com/p/0187459831af