One of the most popular algorithms in processing internet data i.e. web pages is a site ranking algorithm that is intended to decide the importance of web pages by attribution weight value based on any incoming link to this site.Large amounts of internet data can lead to computational load in processing the page ranking algorithm.To take this burden into account, in this article we present a an algorithm for processing site rankings through a distributed system A Hadoop MapReduce framework called MR PageRank. This paper intended for the first analysis of input raw web pages create the page name and its outbound links as key and value of the pair, respectively, as well as the total weight of the hanging nodes and total number of pages. Next, we calculate the probability each page and divide this probability by each outgoing link evenly. Each outgoing weight is mixed and aggregated based on page similarity to new update weight value of each page. Keywords: pagerank, hadoop, mapreduce, Attribution weight...