Configurations for submitting a Hive query to a dedicated queue

When you run a Hive query from Hue or Beeline, HiveServer2 submits the query in YARN in the queue mapped from an end user instead of a hive user. If you run a query while specifying doAs=false, then the queries are submitted and run as a hive user, using impersonation. This is the default behaviour. You can change this behaviour by configuring the queue using "yarn_queue_manager" and "application tag" parameters in YARN.

If you see go to the Job Browser page in Hue and see that the Hive queries are appearing under the default queue, and not the actual user who ran the query, then this means that the queries are not being run in their dedicated queue in YARN. To submit a Hive query in its dedicated queue, you must enable application-tag based scheduling as described in Configuring queue mapping to use the user name from the application tag using Cloudera Manager and Configure Per Queue Properties.