-werners-
Esteemed Contributor III

Databricks does not use vanilla spark.
They added optimizations like the AQE, unity catalog etc.
So looking for the error in the spark source code will not always work (in a lot of cases it will)