The pipeline should store information in the database about datasets which it tried to process but was unable to, along with the error message