Created on 03-15-2019 12:35 PM - edited 08-17-2019 04:43 PM
Hello everyone, is it possible to use Spark with the Hive Warehouse Connector with Ranger while not having Kerberos enabled? On the spark shell I keep getting the error...
FAILED: HiveAccessControlException Permission denied: user [anonymous] does not have [USE] privilege on [default]
... despite configuring every single aspect illustrated in the HDP documentation (image attached).
The user is "anonymous" but I don't understand why because when reading/writing to HDFS my user is established properly. I suspect I have to pass an argument to some Hive Warehouse Connector parameter, but the documentation does not indicate this.
Can someone help?
Bets regards,
Carlos
Created 03-15-2019 02:28 PM
Ok, found it. You just need to add the conf "user" in the following url property:
spark.sql.hive.hiveserver2.jdbc.url=jdbc:hive2://xxxxxx:2181,xxxxxxx:2181,xxxxxxx:2181/;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2-interactive;user=hive
I feel that this case could be better documented :)
Best regards,
Carlos