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?