Omit Impala Catalog




I was able to connect to impala via Cloudera JDBC Driver, but somehow I'm unable to the "CatalogSchemaSwitch=1" property, which I think is the property I'm looking to use.  

Using the Hive JDBC driver, I see everything as it should be, without the default "Hive" Catalog, but I can't seem to do the same with the Impala connection, and in the database explorer I see the deafault "Impala" catalog above all the schemas. 

What i'm looking for is the equivalent of the driver property "Omit Catalog(s)" using DBeaver.


Thanks in advance.



CatalogSchemaSwitch will not help because DataGrip thinks that Impala database has both catalogs and schemas.

Please try to disable Show Intermediate Nodes option in Database View (see screenshot). This should hide all tree levels with only one node. But it may hide more nodes than you wanted :(

Does completion work well for you when you write SQL queries in query consoles?


Hi Liudmila!

Thank you for your response.

Yes, it appears that unselecting "Show Intermediate Nodes" does help, and it shows objects as I'd expect, without any issues so far. 

But to answer your next question, no, autocompletion doesn't seem to be working properly. When I autocomplete the name of a database I get something like select * from "this_is_mydatabase" (with the double quotes) with the next options being "LATERAL" "LATERAL VIEW" etc, without any suggestions of the tables. 

To make the autocompletion suggest the tables properly I have to remove the double quotes from the database name.


Any ideas how to fix that, and insight on this behaviour?



Hello Rafael

Impala uses back ticks ` for quotation, am I right?
I guess we can fix that

Could you please invoke Quick Documentation on the database in Database View (on macos it's F1, on windows and linux I believe it's Ctrl+Q) and send me firsts 5 lines of it (without content)?
I want to make sure that we properly detect the database



Thank you for the assistance vasily chernov and Liudmila Kornilova.

I changed the dialect for the database to HSQLDB, and it seems to work now, though of course some commands like invalidate metadata are detected as errors. 

Here's the print of the lines from the connection that you asked Liudmila Kornilova


Actually, HSQLDB is not what I thought it was. 

I changed back the driver dialect to Hive and everything is working fine still.

It might have been the update I performed to version 201. 

It appears that the problem is indeed fixed either way. 

Thank you again. 


Please sign in to leave a comment.