With the rapid development of Peer-to-Peer (P2P) technology, IPTV applications based on that have received more and more attention from both industry and academia. Several applications using the data-driven mesh-pull architectures raised and gained great success commercially. At present, PPLive system is one of the most popular instances of IPTV applications which attract a large number of users from across the globe. At the same time, however, the dramatic rise in popularity makes it more likely to become a vulnerable target. In this paper, we propose an effective measurement architecture, which is based on the peers nature not only receiving polluted video chunks but also forwarding those to other peers, to measure the video streaming pollution attack and then use a dedicated crawler of PPLive developed by us to evaluate the impact of pollution in P2P live streaming. Specifically, the results show that a single polluter is capable of compromising all the system and its destructiveness is severe.