shan_chandra
Databricks Employee
Databricks Employee

@dmart  -  If it is over partitioned, The easiest way is to solve by collect the paths of the inner partitions recursively, list them and delete them in parallel.