Home > Blockchain >  ADF: Is it possible to gracefully handle issues with a csv to Azure SQL import?
ADF: Is it possible to gracefully handle issues with a csv to Azure SQL import?

Time:09-22

I have a comma separated csv file with double quotes as text identifier. Sometimes the source system export is incorrect, e.g.:

  • A text field includes a double quote causing ADF to think there is one extra column and fail
  • A text field includes the escape character causing ADF to concatenate 2 columns and failing with the error that there is one column less than expected.

The source system vendor is unable to fix this, so these errors will happen every now and then. Is it possible for ADF to just save the whole row into an logfile/logtable and just skip this line?

I am aware of this question, but I can't change the escape character in this case.

Thanks in advance for your answer!

Johan

CodePudding user response:

In ADF, if you are using the Copy Activity, you will use the Fault Tolerance and Enable Logging features to achieve this. When transforming data with a data flow, you will use the "Error row handling" feature to achieve it.

  • Related