Hue使用Oozie执行Spark2任务

Oozie共享库添加Spark2

  1. Oozie用户在/user/oozie/share/lib/lib_2019102914381创建spark2目录

     sudo -u oozie hdfs dfs -mkdir /user/oozie/share/lib/lib_20180605143536/spark2
    

  2. 复制jar包到spark2

     cd /usr/hdp/3.1.0.0-78/spark2/jars
     sudo -u oozie hdfs dfs -put *.jar /user/oozie/share/lib/lib_2019102914381/spark2
    
     cd /usr/hdp/3.1.0.0-78/oozie/share/lib/spark
     sudo -u oozie hdfs dfs -put oozie-sharelib-spark-4.3.1.3.1.0.0-78.jar /user/oozie/share/lib/lib_2019102914381/spark2
    
  3. 修改目录权限

     sudo -u hdfs hdfs dfs -chmod -R 775 /user/oozie/share/lib/lib_20180605143536/spark2
    
  4. 更新Oozie的share-lib

     su oozie
     oozie admin -oozie http://10.47.123.40:11000/oozie -sharelibupdate
    
  5. 查看Oozie的share-lib

     oozie admin -oozie http://10.47.123.40:11000/oozie -shareliblist
    
  6. 创建spark任务

    点击Files,选择/user/oozie/share/lib/lib_2019102914381/spark2,点击选择此文件夹

    Jar/py name填写spark-examples_2.11-2.3.2.3.1.0.0-78.jar

    填写Main class:org.apache.spark.example.SparkPi

    Spark Master: yarn

    Mode: client

    oozie.action.sharelib.for.spark: spark2

    保存并执行

results matching ""

    No results matching ""