However for the Hive Plug in, in the Ranger UI the Autocomplete does not work. Which user handles the lookup for databases and tables for the Hive Plug in, in the Ranger UI? Note, the Hive Plug in, and the authorization works, but the autocomplete of database and tables and columns does not work....
Is it the Ranger repository config user the one that does the lookup for tables and columns to get the autocomplete to work in Ranger UI?
Is the Ranger repository config user only a database user?
What if someone used cluster prefixes for non-local system users upon creation of cluster e.g. <cluster>-ambari-qa and <cluster<-hive? Should the Ranger repository config user still be "hive" as it is not an actual system account BUT the Policy User for Hive be <cluster>-ambari-qa? Trying to make sense of all of this and why the Hive autocomplete does not work.
Lookup is handled by the user which is configured in the Hive Repository in Ranger UI, Also check that user has the policy to do the hive query. It has to be hadoop user for the user to get authorized. In secure cluster it has to be principal with password from kdc i.e. hive@EXAMPLE.COM
Resolution: I used rangeradmin user instead... that existed and that worked.. so I will make a note that clusters with non local users and with cluster prefixes will have this issue...
The Lookup is done by the Repository Configuration User. This user must be created separated as a local or AD user in a non secured cluster, and given a password. Eg. rangeradmin. In a secured cluster created a user principal with a password. A key tab does not work. Use that principal as the configuration user. e.g. rangeradmin@hostname
Please also note that if this user creation step is done AFTER enabling hive plugin (or ANY plugins) from ambari (i.e. the repo is already created and exists in ranger), making changes in Ambari will NOT work, so repo config user and password NEED to be updated in Ranger Admin UI directly. This is because once the repo is created in ranger, Ambari won't try to update it.