Why are column names with __ invalid?

  • 25 February 2022
  • 4 replies


I have a data model in snowflake that I am trying to import into GoodData. The tables have a few columns that have names like __companyId (two underscores in front of the word comapnyId)

When I try to import it in the UI I get a message “Connection failed. Unknown error” - but if I look in the logs I can see the error:

com.gooddata.tiger.metadata.service.ldm.model.GenerateLdmNamingException, message=Column name is not aligned with naming conventions name=__COMPANYID separator=__ numOfParts=2 expectedNumOfParts=3

I will rename the columns in my data so I can move on, but I would like to know why “__” is invalid in a column name. We use that type of column name quite a lot throughout our data to signify extra data that we added to our data model, so we would like to understand the reasoning behind that.


Best answer by jacek 31 March 2022, 22:15

View original

4 replies

Userlevel 3

Hello Chanan,

Please, are you using any prefixes before the double underscore as well? Is this affecting only certain type of columns - such as connection points, attributes...

No, the columns in question all start with __ - such as __COMPANYID

Userlevel 3

Thank you for the clarification, appreciated. I am currently checking this internally with our Engineering team and I will get back to you as soon as possible.

Userlevel 2

Hi Chanan,

unfortunately “__” is the default separator used by our backend to process naming conventions.

However, configuration of the separator is a part of the scan dialog in LDM Modeler. You can delete “__” and scan should work properly.