数据库
首页 > 数据库> > spark sql 读取kudu表向sqlserver数据库中插入70万条数据

spark sql 读取kudu表向sqlserver数据库中插入70万条数据

作者:互联网

spark sql 读取kudu表向sqlserver数据库中插入70万条数据

1.废话不多说。直接上代码。

import java.util.Properties

import org.apache.spark.SparkConf
import org.apache.spark.sql.SparkSession

object KuduToSqlserver {
  val kuduMasters = "cdh-5:xx,cdh-6:xx"
   //TODO 1:定义kudu表
  val kudutableName = "impala::dw_etl.zxjk_day_etl_flow_rg_kudu"

  def main(args: Array[String]): Unit = {
    val conf = new SparkConf()
    conf.setAppName("SparkKudu")
    conf.setMaster("local[*]")
    val option = Map("kudu.master" -> kuduMasters, "kudu.table" -> kudutableName)
    val sparkSession = SparkSession.builder()
      .config(conf)
      .config("hive.metastore.uris", "thrift://cdh-2:9083")
      .config("hive.metastore.warehouse.dir", "hdfs://http://cdh-3:3306/user/hive/warehouse")
      .getOrCreate()
    // TODO: 读取kudu
    sparkSession.read.format("org.apache.kudu.spark.kudu")
      .options(Map("kudu.master" -> kuduMasters, "kudu.table" -> kudutableName)).load.createOrReplaceTempView("tmp_kudu_table")
    val result = sparkSession.sql(
      """
        |select pscode,outputcode,pollutant_code,region_code,province,city,psname,outputname,case pollutant_code when '001' then '颗粒物'when '001' then '二氧化硫'when '003' then '氮氧化物'end pollutantname,
        |focusindustrytype_new,monitortime,reviseflow, updateflow,qx_flow,reason
        |from tmp_kudu_table
        |""".stripMargin)
    println("读取成功")
    val prop = new Properties()
    prop.setProperty("driver", "com.microsoft.sqlserver.jdbc.SQLServerDriver")
    prop.setProperty("user", "你的登录名")
    prop.setProperty("password", "密码")
    result.write.mode("append")
      .jdbc("jdbc:sqlserver://ip:1433;DatabaseName=AMDB_DataCleanDB", "[dbo].[tb_day11]", prop)
    println("写入成功")
    sparkSession.close()
  }
}

总结,本次推入数据大约70多万条,总耗时8分钟,各位如有更快更好的方案,大家一起来讨论,共赏,谢谢大家。

标签:val,sparkSession,表向,sqlserver,prop,70,kudu,spark
来源: https://blog.csdn.net/weixin_49739324/article/details/118784917