Not sure why, but I just have a case where it takes 900s to receive a 260.000 rows table. This should be _much_ faster.