Flickr and every other organization these days are using and managing trillions of big data. The research’s focal point is about analyzing how efficiently video data such as YouTube videos can be managed for storing, and how well the clusters can be formed to optimize content based storage for content based retrieval. The focus lies on cost effectiveness and how timely only relevant video data can be analyzed, stored and then retrieved for use. IBM’s Hadoop and Google’s MapReduce have proved to be very flexible and efficient contributors for big data. The problem however is to efficiently store video data without enough information about clusters or type of groups the data can fall into. Since millions of videos are uploaded every second.