DP-700 Microsoft Fabric Data Engineer Practice Exam

Using incremental CDC changes with sink upserts keeps the destination table in sync by applying only the emitted deltas as they arrive. CDC gives a stream of changes—inserts, updates, and deletes—since the last checkpoint. When you apply these changes with upserts to the sink, each change is merged by the record key: inserts add new rows, updates modify existing rows, and deletes remove rows. This preserves the latest state of the data efficiently, without reprocessing the entire table, and it handles all change types in real time or near real time.

Dropping and reloading the entire table on every change is extremely wasteful and slow, redoing all data instead of just the delta. Rebuilding the dataset from scratch for each change has the same problem. Only deleting changes without inserts would lose new data and updates, leaving the sink inaccurate.