Do not input private or sensitive data. View Qlik Privacy & Cookie Policy.
Skip to main content

Announcements
Qlik Connect 2026 Agenda Now Available: Explore Sessions
cancel
Showing results for 
Search instead for 
Did you mean: 
Timothee_Hamon
Partner - Contributor II
Partner - Contributor II

Oracle ODBC connection via gateway --> data discrepancies

Hello,

During our migration to Qlik Cloud, we are facing an issue with Oracle ODBC connection when we are comparing result between on premise and saas.

We used the last version of gateway, but number are consired as text in Qlik Cloud.

 

Can you help us?

Does anyone face the same issue?

 

Best regards,

Timothée H.

 

 

Labels (2)
1 Solution

Accepted Solutions
Timothee_Hamon
Partner - Contributor II
Partner - Contributor II
Author

Hello,

 

The discrepancies came from the type of the column in the source database.

Even if it has only number in it, it is defined as varchar() in the oracle database.

Since we cannot change the column type in the database, we choose to update our extracting script by using Num#() when needed (column only containing number but declared as varchar()).

 

Why is it different between on-premise and SaaS?
The answer is: we were not using the odbcconnector package from Qlik but the ODBC driver from microsoft server. And by using the driver, when extracting the data from the database, Qlik checked the type of data inside the column and no the type of the column.

With the gateway, and especially the bulk reader option, Qlik do not check the data line by line, but just extract the column using the defined type by the database.

 

Best regards,

Timothée HAMON

View solution in original post

5 Replies
NadiaB
Support
Support

Hi @Timothee_Hamon 

By any chance could you share what is the field data_type on the Oracle DB? Is it a numeric data type or is it a string data type ?

Could you share the specific data type and length if applies ?

What is the max value (largest number) on that field. 

What is the version in Qlik Sense on prem (including any patch)?

Kind Regards. 

 

Don't forget to mark as "Solution Accepted" the comment that resolves the question/issue. #ngm
Timothee_Hamon
Partner - Contributor II
Partner - Contributor II
Author

Hello,

 

The discrepancies came from the type of the column in the source database.

Even if it has only number in it, it is defined as varchar() in the oracle database.

Since we cannot change the column type in the database, we choose to update our extracting script by using Num#() when needed (column only containing number but declared as varchar()).

 

Why is it different between on-premise and SaaS?
The answer is: we were not using the odbcconnector package from Qlik but the ODBC driver from microsoft server. And by using the driver, when extracting the data from the database, Qlik checked the type of data inside the column and no the type of the column.

With the gateway, and especially the bulk reader option, Qlik do not check the data line by line, but just extract the column using the defined type by the database.

 

Best regards,

Timothée HAMON

Raju_6952
Creator III
Creator III

Hi @Timothee_Hamon 

Are you able to fix this issue in saas?

We are also facing the same problem with the hana odbc connection setup through data gateway.

Please kindly let me know if anything needs to checked to fix this issue.

Thanks in advance.

Regards,

Raju

Ia there a way to fix this 

Raju_6952
Creator III
Creator III

Hi @NadiaB ,

Is there any Qlik Bug created for this reported issue?, As we are using the latest version of the data gateway service 1.7.0 which was released on Jan\14\2025 and still facing the above mentioned issues.

Regards,

Siva Prasad

Timothee_Hamon
Partner - Contributor II
Partner - Contributor II
Author

Hello,

Sadly as formentionned, we had to use function like Num#() to enforce type when needed (and it was tedious).

Best regards,
Timothee H.