Hadoop生态圈-Flume的组件之自定义拦截器（interceptor）

　　　　　　　　　　　　　　　　Hadoop生态圈-Flume的组件之自定义拦截器（interceptor）

　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　作者：尹正杰

版权声明：原创作品，谢绝转载！否则将追究法律责任。

　　本篇博客只是举例了一个自定义拦截器的方法，测试字节传输速度。

1>.自定义interceptor方法

 /*
 @author :yinzhengjie
 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E7%94%9F%E6%80%81%E5%9C%88/
 EMAIL:y1053419035@qq.com
 */
 package cn.org.yinzhengjie.interceptor;

 import org.apache.flume.Context;
 import org.apache.flume.Event;
 import org.apache.flume.interceptor.Interceptor;

 import java.util.List;

 /**
  * 设置限速拦截器
  * <p>
  * 当 字节/时间，即同一时刻，如果进入的字节过多
  * 则休眠一会
  */
 public class MyInterceptor implements Interceptor {

     private int speed;

     //构造
     private MyInterceptor(int speed) {
         this.speed = speed;
     }

     //do nothing
     public void initialize() {

     }

     /**
      * 1、拿出上一个event的时间，和当前时间进行相减，得出上一个event的时间间隔
      * 2、得到上一个event的body字节数
      * 3、相除得到上一个event的速度，并在此event中先进行停留，再返回event
      *
      * @param event
      * @return
      */

     long lastTime = -1;
     long lastBodySize = 0;

     public Event intercept(Event event) {

         byte[] body = event.getBody();
         int len = body.length;

         long current = System.nanoTime();

         //第一个event
         if (lastTime == -1) {
             lastTime = current;
             lastBodySize = len;
         }

         //非第一个event
         else {
             //计算上一个event停留的时间
             long interval = current - lastTime;
             System.out.println("=========================" + current + "/" + lastTime + "/" + interval + "=========================");
             //上一个event的速度
             int now_speed = (int) ((double) lastBodySize / interval * 1000);
             if (now_speed > speed) {
                 System.out.println("=========================" + now_speed + "=========================");
                 //计算需要停留多少秒 线程休眠，时间 = shouldTime - interval
                 try {
                     Thread.sleep((lastBodySize / speed) * 1000 - interval);
                 } catch (InterruptedException e) {
                     e.printStackTrace();
                 }
             }
             lastBodySize = len;
             lastTime = System.currentTimeMillis();

         }
         return event;

     }

     //迭代List<Event>，将所有Event交给intercept(Event)进行处理
     public List<Event> intercept(List<Event> events) {
         for (Event event : events) {
             intercept(event);
         }
         return events;
     }

     //do nothing
     public void close() {

     }

     public static class Builder implements Interceptor.Builder {

         private int speed;

         public void configure(Context context) {
             speed = context.getInteger(Constants.SPEED, Constants.DEFAULT_SPEED);

         }

         public Interceptor build() {
             return new MyInterceptor(speed);
         }
     }

     public static class Constants {
         public static String SPEED = "speed";
         public static int DEFAULT_SPEED = 1;

     }
 }

2>.打包并将其发送到 /soft/flume/lib下

[yinzhengjie@s101 ~]$ cd /soft/flume/lib/
[yinzhengjie@s101 lib]$
[yinzhengjie@s101 lib]$ ll | grep MyFlume
-rw-r--r--  1 yinzhengjie yinzhengjie    5231 Jun 20 18:53 MyFlume-1.0-SNAPSHOT.jar
[yinzhengjie@s101 lib]$
[yinzhengjie@s101 lib]$ rm -rf MyFlume-1.0-SNAPSHOT.jar
[yinzhengjie@s101 lib]$
[yinzhengjie@s101 lib]$ rz

[yinzhengjie@s101 lib]$
[yinzhengjie@s101 lib]$ ll | grep MyFlume
-rw-r--r--  1 yinzhengjie yinzhengjie    8667 Jun 20 21:02 MyFlume-1.0-SNAPSHOT.jar
[yinzhengjie@s101 lib]$
[yinzhengjie@s101 lib]$ 

3>.编写agent的配置文件

[yinzhengjie@s101 ~]$ more /soft/flume/conf/yinzhengjie_myInterceptor.conf
# Name the components on this agent
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# 定义源: seq
a1.sources.r1.type = seq
# 定义一次RPC产生的批次数量
a1.sources.r1.batchSize = 

# 指定添加拦截器
a1.sources.r1.interceptors = i1
a1.sources.r1.interceptors.i1.type = cn.org.yinzhengjie.interceptor.MyInterceptor$Builder
a1.sources.r1.interceptors.i1.speed = 

# Describe the sink
a1.sinks.k1.type = logger

# Use a channel which buffers events in memory
a1.channels.c1.type = memory
a1.channels.c1.capacity =
a1.channels.c1.transactionCapacity = 

# Bind the source and sink to the channel
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1
[yinzhengjie@s101 ~]$ 

4>.启动flume并测试

[yinzhengjie@s101 ~]$ flume-ng agent -f /soft/flume/conf/yinzhengjie_myInterceptor.conf -n a1

　　下图是运行agent部分的输出内容