I appreciate you answer, i dont know the jira number but the case is solved by upgrading to hdp 3.1.5. This is the link where you can find "descriptions of new features". Here they mention that hive and spark can interact in an easy way due to the fact that now they are using the same hive metastore. Right now we are using hdp 3.1.4, and when we want to write or read hive objects we have to use the hive warehouse connector, which, sometimes is too slow.   https://docs.cloudera.com/HDPDocuments/HDP3/HDP-3.1.5/release-notes/hdp-release-notes.pdf   Apache component     Hive Feature areas, benefits, and links to further information     Core Capabilities         • Support single catalog for Hive and Spark using the Hive Metastore (HMS) translation layer.   Regards,   ... View more