@Teddy Brewski Below are the properties which control the logs and other files written to /tmp/<username> folder. <property> <name>hive.exec.scratchdir</name> <value>/tmp/hive</value> </property> <property> <name>hive.exec.local.scratchdir</name> <value>/var/log/hadoop/hive/tmp/${user.name}</value> </property> <property> <name>hive.downloaded.resources.dir</name> <value>/var/log/hadoop/hive/tmp/hive/${hive.session.id}_resources</value> </property> <property> <name>hive.server2.logging.operation.log.location</name> <value>/var/log/hadoop/hive/tmp/operations_logs</value> </property> You can add/modify under Ambari -> Hive configs. ... View more

Thank you @Sandeep More That seems to be the one! We're running latest HDP (2.6.1) which comes with Knox 0.12.0. Going to try to convince support to release 0.13.0 RPMs. ... View more

Thank you very much @Santhosh B Gowda -- that was it! ... View more

Running each topology on its own Gateway instance is fine, but it's not necessary. You can use a single Knox Gateway instance and simply create a separate topology per-AD. Say you have 2 topologies, ad1 and ad2, then you can connect using: https://knox-host:8443/gateway/ad1/<service>/. https://knox-host:8443/gateway/ad2/<service>/. ... View more