SAP HANA: "DB GroupBy" fails - when no aggregation

SAP HANA has columns technical names which is German abbreviations.
But each column also has a “pretty” name associated.

Consider the system view: SYS.M_CONNECTIONS

image

Last year?, KNIME made a super feature, to show the pretty names instead of the technical.

Pretty names shows up in nodes like, “DB Column Filter”

And query then looks like:

Pretty names is also shown in “DB GroupBy”.

But creates a faulty SQL

If one ads any aggregation, then it works

The current fix, is to do something manually in a “DB Query Reader” node.

Our users often use the “DB GroupBy” node without aggregation, to get all columns combinations in a view for pre-analysis or looping purpose.

Please have a look at this issue

Hello @tescnovonesis ,

thanks a lot for the detailed problem description. This looks like a problem in the query flattener.
Can you please go to the Advanced tab of the DB Connector node and disable the “Flatten sub-queries where possible” option:


and execute the workflow again.

Thanks
Tobias

4 Likes

Yes, that worked.

Remove tick in “DB Connector” and DB GroupBy Works

1 Like

Thanks for the fast response. I have created a ticket for the problem (internal number AP-23850) and get back to you as soon as I know when we can fix this.
Bye
Tobias

4 Likes

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.

Hello @tescnovonesis ,

the problem will be fixed with the 5.3.3 and 5.4.1 bug fix releases of the KNIME Analytics Platform.

Bye
Tobias

3 Likes

Internal ticket ID: AP-23850
Summary: Query flattener creates invalid statement for SAP Hana if no aggregation column is used in the GroupBy node
Fix version(s): 5.3.4, 5.4.1