Hubert-Dudek
Databricks MVP

What about forEachBatch and then MERGE?

Alternatively, run another process that will clean updates using the window function, as you said.


My blog: https://databrickster.medium.com/