This paper focuses on one of the most used short video servers on the Internet, it is YouTube. YouTube has the third rank of the internet sites related to its traffic transactions. In this work, firstly, we study the effect of the huge numbers of viewers, hits, users and files on the video sharing servers' behaviours by analyzing some recent statistics about YouTube. Then, we propose an optimization for file allocation procedures in general and then we apply the algorithm to some examples of YouTube videos. This solution improves the number of hits related to those kinds of servers over the Internet. Finally, we try to optimize the revenue from the file allocation and propose a hybrid solution for the file hosting or server caching systems.