Salesforce Certified Data 360 Consultant free question 3 / 10 · Domain 3: Data Source Connection and Ingestion
An incremental batch data transform has a filter that removes records whose status is Closed from the output. A source record changes from Open to Closed, and the next run is also incremental. Which result matches the Help article's explanation of data drift?
- The record stays in the output and still remains after a full run until you fix the filter
- The record stays in the output and remains until you delete it by hand
- The record is removed from the output at the time of that incremental run
- The record stays in the output and is removed by the next full run
Show the answer and explanation
Answer: D
Why this is correct
The article says that in incremental runs, data drift can leave changed records in the output without being filtered or replaced. When you filter by value to exclude records, a record updated to a value that now meets the exclusion condition is not removed by an incremental run. It is removed by the next full run. To correct data drift, a full run is performed every 24 hours.
Why the other options are wrong
"The record is removed from the output at the time of that incremental run" … An incremental run does not remove the record even though its changed value now meets the exclusion condition. It is removed by the next full run.
"The record stays in the output and remains until you delete it by hand" … Staying in the output is right, but it does not remain until someone deletes it by hand. A full run corrects the data drift.
"The record stays in the output and still remains after a full run until you fix the filter" … Staying in the output is right, but the article says a record that should have been filtered out is removed by the next full run. You do not have to fix the filter first.
Another kind of drift
If the value of a column mapped to the primary key changes, the generated primary key differs from the original, so the updated record is treated as a new record. The output of that incremental run then holds both the original-key record and the new-key record until the next full run.