Looks like this was caused by Ambari bug. The below Hive config change fixed the issue for me
In Ambari > Hive > Configs > Advanced webhcat-site > templeton.hive.properties
change the port from 9933 to 9083
restart Hive
Also, if you get “IOException: Failed to replace a bad datanode on the existing pipeline…” when creating table
In Ambari > Hive > Configs > General > Block replication
Change 3 to 1
