Experiences on a Video Management Cloud Based on Hadoop

Streaming video data grows rapidly with the massive deployment of surveillance cameras, transportation monitoring and other purposes. Therefore how to deal with the increasing scale of streaming data, for example video data storage, video playing and manipulation are challenging due to the big volume of the video data. This paper presents a massive video data management system which based on Hadoop distributed cloud platform with Linux clusters. We build a video management cloud with J2EE, Red5 streaming technology. A user-friendly interface for managing the streaming media data is implemented. It allows easy expansion of storage capacity. We have evaluated the video cloud for extensibility, performance, and shows that it can satisfy the streaming media data storage management requirements.