Why does Spark crossJoin take so long for a tiny dataframe?

Question

I\'m trying to do the following crossJoin on two dataframes with 5 rows each, but Spark spawns 40000 tasks on my machine and it took 30 seconds to achieve the task. Any idea