Oracle wrong data type mapping

Hello !

I think I probably found a bug with the nodes “DB Connector (Labs)” or “Oracle Connector (Labs)” with the data type mapping.

In the case of the Oracle SQL data type “DATE” the mapping in those nodes is done as follow: “=> LocalDate => Local Date”.

As we can read in the Oracle doc, the DATE data type format contains information about the date and the time

As the mapping is done to a Local Date, the information about the time is lost and it is not possible to select a different type in the mapping.

If I use the classic node “Database Connector” it’s working normally.


Hello! I confirm I have the same behaviour and the same problem. Difficult to analyse our timestamped logs in this context.

thanks for reporting this problem. We will look into it and solve it as soon as possible. The type mapping framework uses the JDBC types as database types. According to the Oracle documentation is the Oracle type Date mapped to the JDBC type timestamp which is correctly mapped to Local Date Time. However it seems that this is not the case. Since the JDBC driver returns Date as JDBC type which is mapped to Local Date. We will investigate further and fix the problem.

we have identified and fixed the problem. The fix will be available with the next KNIME bug fix release. Since Oracle only supports Date and Timestamp which both return the JDBCType Timestamp we will allow user to map from timestamp to LocalDate and LocalDateTime when they are connected to an Oracle db. The default database types suggested by the DB Table Creator node for LocalDate and LocalDateTime cells is now DATE. For LocalDateTime the default type is still Timestamp. Let us know if you have found any other issues or if you have any suggestions. Thanks for testing the new database framework.

1 Like

Thanks @tobias.koetter for this very good news. Is there already a planned date for this bug fix release?

Hi there!

Check this topic regarding release:


1 Like

Thanks - I’ll also wait for end January then!

I just donwloaded and updated Knime Analytics Platform to version 3.7.1.
The bug is indeed corrected.
Thanks a lot for taking into account our findings!