视频转码是一种计算密集型和数据密集型的工作，需要消耗大量的计算机资源，Hadoop MapReduce是一种软件模型，支持大规模数据集的分布式处理．．利用已有开源视频转码软件ffmpeg建立一个基于MapReduce框架的Hadoop集群转码系统，提高视频转码的速度．%Video transcoding is a both compute-intensive and data-intensive task. It can consume a large amount of computer resources. Mapreduce is a software framework which is designed to support for distributed processing of large data sets. This paper aims to use open source video transcoding software （ffmpeg） to create a Hadoop cluster to improve the speed of video transcoding.
展开▼
