为什么采用直连(createDirectStream)的方式,主要有以下几个原因:
1.createDirectStream的方式从Kafka集群中读取数据,并且在Spark Streaming系统里面维护偏移量相关的信息,实现零数据丢失,保证不重复消费,比createStream更高效;
但是采用直连(createDirectStream)的方式有一个缺点,就是不再向zookeeper中更新offset信息。2.创建的DStream的rdd的partition做到了和Kafka中topic的partition一一对应。
因此,在采用直连的方式消费kafka中的数据的时候,大体思路是首先获取保存在zookeeper中的偏移量信息,根据偏移量信息去创建stream,消费数据后再把当前的偏移量写入zookeeper中。在创建stream时需要考虑以下几点:
1.zookeeper中没有偏移量信息,此时按照自定义的kafka参数的配置创建stream;
2.zookeeper中保存了偏移量信息,但由于各种原因kafka清理掉了该处偏移量的数据,此时需要对偏移量进行修正,否则在运行时会出现偏移量越界的异常。 解决方法是调用spark-streaming-kafka API 中 KafkaCluster这个类中的方法获取broker中实际的最大最小偏移量,和zookeeper中偏移量进行对比来修正偏移量信息。在2.0以前的版本中KafkaCluster这个类是private权限的,需要把它拷贝到项目里使用。2.0以后的版本中修改KafkaCluster的权限为public,可以尽情调用了。
为了方便调用,本人在使用时写了一个KafkaHelper的类,将创建stream和更新zookeeper中offset的代码封装了起来,代码如下:
import kafka.common.TopicAndPartition
import kafka.message.MessageAndMetadata
import kafka.serializer.StringDecoder
import kafka.utils.{ZKGroupTopicDirs, ZkUtils}
import org.I0Itec.zkclient.ZkClient
import org.apache.spark.SparkException
import org.apache.spark.streaming.StreamingContext
import org.apache.spark.streaming.dstream.InputDStream
import org.apache.spark.streaming.kafka.{KafkaCluster, KafkaUtils, OffsetRange}
import org.apache.spark.streaming.kafka.KafkaCluster.Err
/**
* KafkaHelper类提供两个共有方法,一个用来创建direct方式的DStream,另一个用来更新zookeeper中的消费偏移量
* @param kafkaPrams kafka配置参数
* @param zkQuorum zookeeper列表
* @param group 消费组
* @param topic 消费主题
*/
class KafkaHelper(kafkaPrams:Map[String,String],zkQuorum:String,group:String,topic:String) extends Serializable{
private val kc = new KafkaCluster(kafkaPrams)
private val zkClient = new ZkClient(zkQuorum)
private val topics = Set(topic)
/**
* 获取消费组group下的主题topic在zookeeper中的保存路径
* @return
*/
private def getZkPath():String={
val topicDirs = new ZKGroupTopicDirs(group,topic)
val zkPath = topicDirs.consumerOffsetDir
zkPath
}
/**
* 获取偏移量信息
* @param children 分区数
* @param zkPath zookeeper中的topic信息的路径
* @param earlistLeaderOffsets broker中的实际最小偏移量
* @param latestLeaderOffsets broker中的实际最大偏移量
* @return
*/
private def getOffsets(children:Int,zkPath:String,earlistLeaderOffsets:Map[TopicAndPartition, KafkaCluster.LeaderOffset],latestLeaderOffsets: Map[TopicAndPartition, KafkaCluster.LeaderOffset]): Map[TopicAndPartition, Long] = {
var fromOffsets: Map[TopicAndPartition, Long] = Map()
for(i <- 0 until children){
//获取zookeeper记录的分区偏移量
val zkOffset = zkClient.readData[String](s"${zkPath}/${i}").toLong
val tp = TopicAndPartition(topic,i)
//获取broker中实际的最小和最大偏移量
val earlistOffset: Long = earlistLeaderOffsets(tp).offset
val latestOffset: Long = latestLeaderOffsets(tp).offset
//将实际的偏移量和zookeeper记录的偏移量进行对比,如果zookeeper中记录的偏移量在实际的偏移量范围内则使用zookeeper中的偏移量,
//反之,使用实际的broker中的最小偏移量
if(zkOffset>=earlistOffset && zkOffset<=latestOffset) {
fromOffsets += (tp -> zkOffset)
}else{
fromOffsets += (tp -> earlistOffset)
}
}
fromOffsets
}
/**
* 创建DStream
* @param ssc
* @return
*/
def createDirectStream(ssc:StreamingContext):InputDStream[(String, String)]={
//----------------------获取broker中实际偏移量---------------------------------------------
val partitionsE: Either[Err, Set[TopicAndPartition]] = kc.getPartitions(topics)
if(partitionsE.isLeft)
throw new SparkException("get kafka partitions failed:")
val partitions = partitionsE.right.get
val earlistLeaderOffsetsE: Either[Err, Map[TopicAndPartition, KafkaCluster.LeaderOffset]] = kc.getEarliestLeaderOffsets(partitions)
if(earlistLeaderOffsetsE.isLeft)
throw new SparkException("get kafka earlistLeaderOffsets failed:")
val earlistLeaderOffsets: Map[TopicAndPartition, KafkaCluster.LeaderOffset] = earlistLeaderOffsetsE.right.get
val latestLeaderOffsetsE: Either[Err, Map[TopicAndPartition, KafkaCluster.LeaderOffset]] = kc.getLatestLeaderOffsets(partitions)
if(latestLeaderOffsetsE.isLeft)
throw new SparkException("get kafka latestLeaderOffsets failed:")
val latestLeaderOffsets: Map[TopicAndPartition, KafkaCluster.LeaderOffset] = latestLeaderOffsetsE.right.get
//----------------------创建kafkaStream----------------------------------------------------
var kafkaStream:InputDStream[(String, String)]=null
val zkPath: String = getZkPath()
val children = zkClient.countChildren(zkPath)
//根据zookeeper中是否有偏移量数据判断有没有消费过kafka中的数据
if(children > 0){
val fromOffsets:Map[TopicAndPartition, Long] = getOffsets(children,zkPath,earlistLeaderOffsets,latestLeaderOffsets)
val messageHandler = (mmd: MessageAndMetadata[String, String]) => (mmd.topic, mmd.message())
//如果消费过,根据偏移量创建Stream
kafkaStream = KafkaUtils.createDirectStream[String, String, StringDecoder, StringDecoder, (String, String)](
ssc, kafkaPrams, fromOffsets, messageHandler)
}else{
//如果没有消费过,根据kafkaPrams配置信息从最早的数据开始创建Stream
kafkaStream = KafkaUtils.createDirectStream[String, String, StringDecoder, StringDecoder](ssc, kafkaPrams, topics)
}
kafkaStream
}
/**
* 更新zookeeper中的偏移量
* @param offsetRanges
*/
def updateZkOffsets(offsetRanges:Array[OffsetRange])={
val zkPath: String = getZkPath()
for( o <- offsetRanges){
val newZkPath = s"${zkPath}/${o.partition}"
//将该 partition 的 offset 保存到 zookeeper
ZkUtils.updatePersistentPath(zkClient, newZkPath, o.fromOffset.toString)
}
}
}
import org.apache.spark.SparkConf
import org.apache.spark.streaming.dstream.InputDStream
import org.apache.spark.streaming.kafka.{HasOffsetRanges, OffsetRange}
import org.apache.spark.streaming.{Seconds, StreamingContext}
object TestKafkaHelper {
def main(args: Array[String]): Unit = {
if(args.length<5){
println("Usage: ")
System.exit(1)
}
val Array(timeInterval,brokerList,zkQuorum,topic,group) = args
val conf = new SparkConf().setAppName("KafkaDirectStream").setMaster("local[2]")
val ssc = new StreamingContext(conf,Seconds(timeInterval.toInt))
//kafka配置参数
val kafkaParams = Map(
"metadata.broker.list" -> brokerList,
"group.id" -> group,
"auto.offset.reset" -> kafka.api.OffsetRequest.SmallestTimeString
)
val kafkaHelper = new KafkaHelper(kafkaParams,zkQuorum,topic,group)
val kafkaStream: InputDStream[(String, String)] = kafkaHelper.createDirectStream(ssc)
var offsetRanges = Array[OffsetRange]()
kafkaStream.transform( rdd =>{
offsetRanges = rdd.asInstanceOf[HasOffsetRanges].offsetRanges
rdd
}).map( msg => msg._2)
.foreachRDD( rdd => {
rdd.foreachPartition( partition =>{
partition.foreach( record =>{
//处理数据的方法
println(record)
})
})
kafkaHelper.updateZkOffsets(offsetRanges)
})
ssc.start()
ssc.awaitTermination()
ssc.stop()
}
}