Pipeline Copy Activity Column does not Exist in Parquet

[rwhitePWT]

Hi,

I'm having issues with a copy activity I am running. It's a pipeline that has a for each setup that runs through an array of variables to a GET request for an API. The output is a parquet file that goes into a azure dfs storage, where I then copy to a lakehouse table.

When I try and push from the storage to the lakehouse table, I am constantly getting this error. 

Activity failed because an inner activity failed; Inner activity name: Copy data6, Error: ErrorCode=ParquetColumnNotFound,'Type=Microsoft.DataTransfer.Common.Shared.HybridDeliveryException,Message=Column schema.table.date does not exist in Parquet file.,Source=Microsoft.DataTransfer.Richfile.ParquetTransferPlugin,'

The setup for the Copy data6 is set up with File path type: File Path; Format: Parquet; and the mapping has been set.

Can anyone give me some guidance?

[jwinchell40]

@rwhitePWT - Is it possible to post a picture of the pipeline itself and the mapping for the 2nd Copy Step?

If I am following correctly:

- You have an Array Variable defined in your Pipeline which has the details on how to pull data

- The ForEach loops through this Array

- In the Child Definition; you have a Copy Activity making an API call (GET) with the results getting copied to ADLS Gen2

- You have a second copy step that is moving that file from ADLS Gen2 to a Fabric Lakehouse (Error happens here)

Questions:

- Curious as to why you are copying the data to ADLS Gen2 but then bringing it back into Fabric; instead of just writing the API output to the "Files" section of Fabric.  There are definitely legitimate reasons for doing this; just curious.

- Is this happening to the output of all the API files or do some make the round trip just fine?

- For the file written to ADLS Gen2 are all of your date fields set to datetime as the data type vs just Date?  I have run into issues where creating a Lakehouse Table with a "Date" datatype will fail because it needs to be a DateTime.  Might be worth checking out.