Created 05-17-2016 03:23 AM
I understand with ranger security in place users may browse all data in hive via SparkSQL? Is there a work around as a stop gap until official bug is fixed in apache? I am aware of this HCC post. However it does not have a resolution or clarity in terms of gap stop.
Created 05-17-2016 02:16 PM
@Sunile Manjee I am not sure this has anything to do with Ranger. As mentioned in this post , if you are using Spark SQL client it will behave similar to Hive CLI today where HDFS permissions come into the play. There is Spark thriftserver work which is pending https://issues.apache.org/jira/browse/SPARK-8659
Created 05-17-2016 02:16 PM
@Sunile Manjee I am not sure this has anything to do with Ranger. As mentioned in this post , if you are using Spark SQL client it will behave similar to Hive CLI today where HDFS permissions come into the play. There is Spark thriftserver work which is pending https://issues.apache.org/jira/browse/SPARK-8659