The Salesforce Connector uses Bulk Data loads for inserts and upserts. To handle errors that occur during Bulk Data jobs, use the Bulk Data Load Jobs section of Monitoring. 

Salesforce Recipe Execution Errors

Here is an example of error from the recipe execution logs.

The logs state how many records failed. To find the detail use Salesforce's built-in monitoring tools.

Identifying and Fixing Bulk Data Errors

  1. Find the Bulk Data Load job that correlates with your recipe statement
  2. Look at the batches for the Bulk Data Load. Lingk automatically creates batches for large Bulk Data Load jobs. Additionally, you can configure the batch size on the connector.
  3. Identify if the type of error.
    Batch level status errors mean that no rows in the entire batch could be processed (most of the time due to duplicate or missing required fields)Batch level results (row errors) mean that there are data validation issues. Some records may succeed and others fail
    Click "View Result" to download a file containing the row by row results. Find the records that fail and update your recipe accordingly. Many times you will find data that needs to be trimmed, data values that need transformation or picklist values that are not mapped (or unavailable in the target object).
  4. Fix the error and try again.

While testing your recipe, you can write out the results of your statements to a CSV file using the Local File connector for review without submitting to Salesforce.