spark streaming整合flume(常用的poll类型)

在实际生产工作环境中，spark streaming经常和flume或者kafka整合在一起使用，本片文章说下与flume整合过程。

常用的整合方式有两种，一种是push类型，一种是poll类型，在实际分布式生产环境下，我们使用poll类型，也就是由spark streaming向flume拿数据，push类型是flume向streaming 送数据。我们这里只说下poll类型的整合方式。

1，安装flume（不解释，之前文章有详解）

2，将这三个jar包放到安装好的flume的lib目录下。  

（链接：https://pan.baidu.com/s/1_zouri5jOGMdk9J3d7nDsQ 密码：v001）

commons-lang3-3.3.2.jar ， scala-library-2.10.5.jar ， spark-streaming-flume-sink_2.10-1.6.1.jar

3，添加 flume-poll.conf 到flume的conf目录下

# Name the components on this agent
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# source
a1.sources.r1.type = spooldir
a1.sources.r1.spoolDir = /export/data/flume
a1.sources.r1.fileHeader = true

# Describe the sink
a1.sinks.k1.type = org.apache.spark.streaming.flume.sink.SparkSink
a1.sinks.k1.hostname = 192.168.2.201
a1.sinks.k1.port = 8888

# Use a channel which buffers events in memory
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# Bind the source and sink to the channel
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

4，启动flume

 bin/flume-ng agent -n a1 -c conf -f conf/flume-poll.conf 

5,  编写scala spark streaming程序 (wordcount小程序，统计同一字符出现的个数)

package cn.itcast.spark.day5

import java.net.InetSocketAddress

import org.apache.spark.SparkConf
import org.apache.spark.storage.StorageLevel
import org.apache.spark.streaming.flume.FlumeUtils
import org.apache.spark.streaming.{Seconds, StreamingContext}
object FlumePollWordCount {
  def main(args: Array[String]) {
    val conf = new SparkConf().setAppName("FlumePollWordCount").setMaster("local[2]")
    val ssc = new StreamingContext(conf, Seconds(5))
    //从flume中拉取数据(flume的地址)
    val address = Seq(new InetSocketAddress("192.168.2.201", 8888))
    val flumeStream = FlumeUtils.createPollingStream(ssc, address, StorageLevel.MEMORY_AND_DISK)
    val words = flumeStream.flatMap(x => new String(x.event.getBody().array()).split(" ")).map((_,1))
    val results = words.reduceByKey(_+_)
    results.print()
    ssc.start()
    ssc.awaitTermination()
  }
}

6， 启动scala程序

7， 在虚拟机的 /export/data/flume 目录下添加一个文件 word.txt ,里面随便写一下东西

如： a a a a b b b c c d d

8, 查看scala程序后台输出：

9， 完毕

10，集群方式部署启动 （push模式）

flume conf 添加 flume-push.conf 文件

# Name the components on this agent
a1.sources = r1
a1.sinks = k1
a1.channels = c1

# source
a1.sources.r1.type = spooldir
a1.sources.r1.spoolDir = /export/data/flume
a1.sources.r1.fileHeader = true

# Describe the sink
a1.sinks.k1.type = org.apache.spark.streaming.flume.sink.SparkSink
a1.sinks.k1.hostname = 192.168.2.201
a1.sinks.k1.port = 8888

# Use a channel which buffers events in memory
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100

# Bind the source and sink to the channel
a1.sources.r1.channels = c1
a1.sinks.k1.channel = c1

#启动flume
bin/flume-ng agent -n a1 -c conf/ -f conf/flume-push.conf -Dflume.root.logger=WARN,console
#再启动spark-streaming应用程序

package cn.itcast.spark.day5

import org.apache.spark.SparkConf
import org.apache.spark.streaming.flume.FlumeUtils
import org.apache.spark.streaming.{Seconds, StreamingContext}

/**
  * Created by ZX on 2015/6/22.
  */
object FlumePushWordCount {

  def main(args: Array[String]) {
    //集群方式启动
    val host = args(0)
    val port = args(1).toInt
    //LoggerLevels.setStreamingLogLevels()
    val conf = new SparkConf().setAppName("FlumeWordCount")//.setMaster("local[2]")
    val ssc = new StreamingContext(conf, Seconds(5))
    //推送方式: flume向spark发送数据
    val flumeStream = FlumeUtils.createStream(ssc, host, port)
    //flume中的数据通过event.getBody()才能拿到真正的内容
    val words = flumeStream.flatMap(x => new String(x.event.getBody().array()).split(" ")).map((_, 1))

    val results = words.reduceByKey(_ + _)
    results.print()
    ssc.start()
    ssc.awaitTermination()
  }
}

bin/spark-submit --master spark://weekend01:7077 --class cn.itcast.spark.day5.FlumePushWordCount /home/bigdata/streaming-1.0.jar 192.168.2.201 8888

完活 ！