ADF data flow wrongly writes DelimitedText/CSV

Poel van der, RE (Ron) 421 Reputation points


we have a pipeline with data flow. Input are SQL Server tables, output csv.


Values are filled as follows


When using a Copy activity in a pipeline everything works fine. We get a csv, layout correct
But if we use the same dataset as a Sink in a data flow, it keeps on writing Parquet, whatever we change.

Only if we change in Settings of the Sink 'File name options' from Default to Output to single file, and in Optimize tab set Single partitioning on, it will create a file with extension csv.


But the content is still Parquet, some binary format.

What is the issue here?
Does anybody know?


Azure Data Factory
Azure Data Factory
An Azure service for ingesting, preparing, and transforming data at scale.
9,504 questions
{count} votes

1 answer

Sort by: Most helpful
  1. Poel van der, RE (Ron) 421 Reputation points

    HI @MartinJaffer-MSFT

    somebody advised me to delete the Sink and create it again.
    I did and the problem was solved. I am getting a csv with correct content.
    What caused the issue is unclear to me.

    So tip. In these cases first delete and recreate the Sink.

    Thnx for your help.