Skip to main content

Community

Connector Improvement: Give destination tables the same column order as the source

Please sign in to leave a comment.

Comments

3 comments

  • Official comment

    Hi Tomas,

    Thank you for taking to the time to share your idea for making Fivetran better. Can you mention which connector service you are using?

    I'd also like to understand more about why the column order is important? - doesn't the order get controlled in your SQL as you actually use the tables?

    Best regards
    Alison

    Hi Alison,

    This request is for our Oracle database connectors.

    Here's why I think it's important:

    I use query tools like DBeaver or SQL Developer to query the tables. These tools have built in queries for base tables. Yes, I can reorder the columns from the generated SQL in order to get them to match but it's an extra step and another chance to make an error.

    There are times when I'm trying to validate the data between source and destination so I have the 2 app windows side by side. If the Snowflake table had the Fivetran generated columns, e.g., _FIVETRAN_DELETED, at the end of the row I could scroll both windows and verify the data in a simpler way. 

    Another thing is that I 'know' column B is next to column A in the source table so that's where I naturally look in the destination. If the destination has column B far away from column A it just makes things harder.

    The last point is that in .csv files, empty columns don't get created in the destination. If the two structures match it's easier to see which columns didn't make it over...

    You can call me lazy, I won't be offended at all! ;-0)

    Thanks

    -Tom

     

     

     

    We have the same issue from SQL Server on Prem.
    For now columns in destination is in random order.
    We would prefer to have the same order as in the source, without having to make a view.
    If new colums in the source are in the end,  we would like to have new columns in the end as well

    Best Regards Ingrid