Skip to main content
Docs
Feature Requests Onboarding Trust Center Status Page

Community

Destination Improvement: Column Ordering

pavan Kumar kambam User

If source has created a table and when we query the table, the column ordering is like below

id, year, month, day, ..........

but when we query the same table in Destination/Snowflake, id and DATE columns are not displaying in the first.

Can you please implement the column ordering in destination as same as source or customer ordering?

Please sign in to leave a comment.

Comments

9 comments

Date Votes
  • Official comment
    Kevin Kim Fivetranner

    Hi all,

    Thanks for your comments on this thread -- while we don't have a timeline for this feature request, please continue to share your feedback here so we can follow up on this feature.

    I will be sure to share this feedback with our internal team for discussion!

  • pavan Kumar kambam User

    Hi Can you please look into it and update if this is possible and can be doen soon?

  • Marina Autina User

    I want to upvote this feature. We are adding a table with over 600 columns. It's going to make the table a lot more usable if I could rely on order being preserved in this case. Creating a custom view that reorders the columns back to the original form will also take a while since all column names have been auto-adjusted. An option to preserve order would be very helpful. 

  • Roy Pruett User

    I am replicating hundreds, if not thousands, of tables. It would be extremely beneficial for end users for Fivetran to apply the same column ordering in the destination as found in the source.

  • Daniel Upton User
    • Edited

    Agree.   For ERP's with hundreds of columns per table, not guaranteeing column order is a substantial dis-satisfier.  Default functionality should guarantee column order per the data source.  If schema change occurs in a data source, one that effects column order in a table, except to the extent that the destination (in our case Snowflake), prevents it.

  • Admin LER User

    Agree, this creates double work for us and unnecessary steps in the setup to re-create the correct order. 

  • Andrew Taft User

    Completely agree, not maintaining ordinal position is a bug and should be dealt with quickly.

    There are at least two other community posts about this:

    1. https://support.fivetran.com/hc/en-us/community/posts/1500000891061-Other-Please-make-the-column-ordinal-positioning-in-Destination-after-FIVETRAN-sync-match-with-the-Source-Database-column-ordinal-positioning-
    2. https://support.fivetran.com/hc/en-us/community/posts/4420223223831-Other-Handling-ordinal-postition-on-pipelines-between-RDBMS
  • Ilya Pruslin User

    Hello, I have a case when I have to merge a table with identical structure from the 3 different database and there are 132 tables in each database. Due to the Fivetran limitation I cannot use guery  select * from schema_nameA.table _name union all select * from schema_nameB.table _name to merge data from the tables across 3 Snowflake schemas,  

  • kazuya sasaki User

    Fivetran support told me that the column order in the destination table is determined by the order in which the data was extracted, but the logic at the time of extraction is undefined.

    I also understand that columns may be added or deleted while using Fivetran, and it is not realistic to guarantee that the column order of the source and destination will match permanently.

    Therefore, I would like to have the column order of the destination table reproduced with the column order of the source table at the time of the initial synchronization. This will ensure that most column orders will match, except for inconsistencies that occur during operation, making it easier for us, the users, to understand the destination table.

    This also means that inconsistencies in column order caused by adding or deleting columns later on will be tolerated.