You are given the data set of all Google crawled URLs on the Internet -- that's a very large set. Write an algorithm to find the median URL length.

What is the most efficient way to sort a million 32-bit integers?