kafka实现读文件到hbase 并且实现hbase到hive上的映射

首先还是构建一个父类特质方法

package kafkatohbase

import org.apache.hadoop.hbase.client.Put
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords}

trait DataHandler {
  def transform(topic:String,records:ConsumerRecords[String,String]):Array[Put]
}

这次读取5个消息队列所以书写了5个特质子类

package kafkatohbase
import org.apache.hadoop.hbase.client.Put
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords}
import org.apache.kafka.common.TopicPartition

trait EventsAttendeesImpl extends DataHandler {
  override def transform(topic:String,records: ConsumerRecords[String, String]): Array[Put] = {
    val topicPartition = new TopicPartition(topic, 0)
    records.records(topicPartition).toArray.map(r=>{
      val rec=r.asInstanceOf[ConsumerRecord[String,String]]
      val info=rec.value().split(",")
      val put=new Put((info(0)+""+info(1)+info(2)).getBytes)
      put.addColumn("base".getBytes(),"user_id".getBytes(),info(0).getBytes)
      put.addColumn("base".getBytes(),"friend_id".getBytes(),info(1).getBytes)
      put.addColumn("base".getBytes(),"attend_type".getBytes(),info(2).getBytes)
      put
    })
  }
}

package kafkatohbase
import org.apache.hadoop.hbase.client.Put
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords}
import org.apache.kafka.common.TopicPartition

trait EventsImpl extends DataHandler {
    override def transform(topic:String,records: ConsumerRecords[String, String]): Array[Put] = {
      val topicPartition = new TopicPartition(topic, 0)
      records.records(topicPartition).toArray.map(r=>{
        val rec=r.asInstanceOf[ConsumerRecord[String,String]]
        val info=rec.value().split(",",-1)
        val put=new Put((info(0)+""+info(1)).getBytes())
        //event_id,user_id,start_time,city,state,zip,country,lat,lng
        put.addColumn("base".getBytes(),"event_id".getBytes(),info(0).getBytes)
        put.addColumn("base".getBytes(),"user_id".getBytes(),info(1).getBytes)
        put.addColumn("base".getBytes(),"start_time".getBytes(),info(2).getBytes)
        put.addColumn("base".getBytes(),"city".getBytes(),info(3).getBytes)
        put.addColumn("base".getBytes(),"state".getBytes(),info(4).getBytes)
        put.addColumn("base".getBytes(),"zip".getBytes(),info(5).getBytes)
        put.addColumn("base".getBytes(),"country".getBytes(),info(6).getBytes)
        put.addColumn("base".getBytes(),"lat".getBytes(),info(7).getBytes)
        put.addColumn("base".getBytes(),"lng".getBytes(),info(8).getBytes)
        var other=""
        for(i<-9 until info.length){
          other+=info(i) + " "
        }
        put.addColumn("base".getBytes(),"other".getBytes(),other.trim.getBytes)
        put
      })
  }
}

package kafkatohbase
import org.apache.hadoop.hbase.client.Put
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords}
import org.apache.kafka.common.TopicPartition

trait TrainImpl extends DataHandler {
  override def transform(topic:String,records: ConsumerRecords[String, String]): Array[Put] = {
    val topicPartition=new TopicPartition(topic,0)
    records.records(topicPartition).toArray.map(r=>{
      val rec=r.asInstanceOf[ConsumerRecord[String,String]]
      val info=rec.value().split(",")
      (info(0),info(1),info(2),info(3),info(4),info(5))
    }).sortBy(_._4).map(x=>{
      val put=new Put((x._1+""+x._2).getBytes())
      put.addColumn("base".getBytes(),"user_id".getBytes(),x._1.getBytes)
      put.addColumn("base".getBytes(),"event_id".getBytes(),x._2.getBytes)
      put.addColumn("base".getBytes(),"invited".getBytes(),x._3.getBytes)
      put.addColumn("base".getBytes(),"timestamp".getBytes(),x._4.getBytes)
      put.addColumn("base".getBytes(),"interested".getBytes(),x._5.getBytes)
      put.addColumn("base".getBytes(),"not_interested".getBytes(),x._6.getBytes)
      put
    })
  }
}

package kafkatohbase
import org.apache.hadoop.hbase.client.Put
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords}
import org.apache.kafka.common.TopicPartition

trait UserFriendImpl extends DataHandler {
  override def transform(topic:String,records: ConsumerRecords[String, String]): Array[Put] = {
    val topicPartition = new TopicPartition(topic, 0)
    records.records(topicPartition).toArray.map(r=>{
      val rec=r.asInstanceOf[ConsumerRecord[String,String]]
      val info=rec.value().split(",",-1)
      val put=new Put((info(0)+""+info(1)).getBytes())
      put.addColumn("base".getBytes(),"user_id".getBytes(),info(0).getBytes)
      put.addColumn("base".getBytes(),"friend_id".getBytes(),info(1).getBytes)
      put
    })
  }
}

package kafkatohbase
import org.apache.hadoop.hbase.client.Put
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords}
import org.apache.kafka.common.TopicPartition

trait UserImpl extends DataHandler {
  override def transform(topic:String,records: ConsumerRecords[String, String]): Array[Put] = {
    val topicPartition = new TopicPartition(topic, 0)
    records.records(topicPartition).toArray.map(r=>{
      val rec=r.asInstanceOf[ConsumerRecord[String,String]]
      val info=rec.value().split(",",-1)
      val put=new Put(info(0).getBytes)
      put.addColumn("base".getBytes(),"user_id".getBytes(),info(0).getBytes)
      put.addColumn("base".getBytes(),"locale".getBytes(),info(1).getBytes)
      put.addColumn("base".getBytes(),"birthyear".getBytes(),info(2).getBytes)
      put.addColumn("base".getBytes(),"gender".getBytes(),info(3).getBytes)
      put.addColumn("base".getBytes(),"joinedAt".getBytes(),info(4).getBytes)
      put.addColumn("base".getBytes(),"location".getBytes(),info(5).getBytes)
      put.addColumn("base".getBytes(),"timezone".getBytes(),info(6).getBytes)
      put
    })
  }
  }

实现读取方法

package kafkatohbase

import java.time.Duration
import java.util

import com.fenbi.mydh.streamhandler.Params
import org.apache.hadoop.hbase.{HBaseConfiguration, TableName}
import org.apache.hadoop.hbase.client.{ConnectionFactory, Put}
import org.apache.kafka.clients.consumer.{ConsumerRecord, ConsumerRecords, KafkaConsumer}


class ReadKafkaToHbase {
  self:DataHandler=>
  def job(ip:String,group:String,topic:String,hbaseTable:String)={
    val hbaseconf=HBaseConfiguration.create()
    hbaseconf.set("hbase.zookeeper.quorum",ip+":2181")
    //读取kafka消息
    val conn= ConnectionFactory.createConnection(hbaseconf)
    val param=new Params
    param.IP=ip+":9092"
    param.GROUPID=group
    val consumer=new KafkaConsumer[String,String](param.getReadKafkaParam)
    consumer.subscribe(util.Arrays.asList(topic))
    while (true){
      val rds:ConsumerRecords[String,String]=consumer.poll(Duration.ofSeconds(5))
      //转换数据获得Array[Put]
      val puts:Array[Put]=transform(topic,rds)
      //将scalaArray【put}转为java ArrayList
      val lst=new util.ArrayList[Put]()
      puts.foreach(p=>lst.add(p))
      val table=conn.getTable(TableName.valueOf(hbaseTable))
      //将数据发送到hbase
      table.put(lst)
      println("..........................................")
    }
  }
}

注意我这里的5个文件已经读取到kafka中了

 然后启动你的hbase

后面还需要用到hive 所以一起启动吧

habse shell 启动编程

首先创建名空间

create_namespace ’exp‘

创建需要的5张表

 然后返回idea 书写test文件

package kafkatohbase

object Test1 {
  def main(args: Array[String]): Unit = {

//    (new ReadKafkaToHbase() with EventsAttendeesImpl)
//      .job("192.168.80.181","ea08","event_attendee","exp:hbase_eventAttendees") //11245008
//    (new ReadKafkaToHbase() with EventsImpl)
//      .job("192.168.80.181","et05","events","exp:hbase_events")  //3137972
//    (new ReadKafkaToHbase() with UserImpl)
//      .job("192.168.80.181","us01","users","exp:hbase_users")  //38209
//    (new ReadKafkaToHbase() with UserFriendImpl)
//      .job("192.168.80.181","uff07","friends","exp:hbase_userFriend")  //30386387
//    (new ReadKafkaToHbase() with TrainImpl)
//      .job("192.168.80.181","tr01","trains","exp:hbase_train") //15220
  }
}

最好是一张一张表的读 然后查看数据是否准确 hbase可以根据列簇去重

全部完成后打开zeppelin 

创建一个属于你的工程

 我这里创建的名字为hbasetest的工程

-- create database hbasetest  创建数据库
--use hbasetest 使用数据库
--表一
-- create external table hbasetest.hbase_events(
-- id string,
-- event_id string,
-- user_id string,
-- start_time string,
-- city string,
-- state string,
-- zip string,
-- country string,
-- lat string,
-- lng string,
-- other string
-- )
-- row format serde 'org.apache.hadoop.hive.hbase.HBaseSerDe'
-- stored by 'org.apache.hadoop.hive.hbase.HBaseStorageHandler'
-- with serdeproperties(
-- "hbase.columns.mapping"=":key,base:event_id,base:user_id,base:start_time,base:city,base:state,base:zip,base:country,base:lat,base:lng,base:other"
-- )
-- tblproperties(
-- "hbase.table.name"="exp:hbase_events"
-- )
--表二
create external table hbasetest.hbase_eventAttendees(
id string,
user_id string,
friend_id string,
attend_type string
)
row format serde 'org.apache.hadoop.hive.hbase.HBaseSerDe'
stored by 'org.apache.hadoop.hive.hbase.HBaseStorageHandler'
with serdeproperties(
"hbase.columns.mapping"=":key,base:user_id,base:friend_id,base:attend_type"
)
tblproperties(
"hbase.table.name"="exp:hbase_eventAttendees"
)
-- select * from hbasetest.hbase_events  查询一下数据是否存在

 数据存在打完收工