I have a doubt, when we broadcast a dataframe, copies of broadcasted dataframe is sent to each Executors.
So , when Spark evict these copies from each Executotrs ?
