关键词: 分布式文件系统, 数据容错, MooseFS, 纠错码

Abstract: In the cloud storage technology, data fault-tolerance of cloud storage system is very important, which is directly related to the availability of the whole system. The current majority of distributed storage system through multiple transcripts to ensure availability of data, however, many copies storage makes double increase data storage space. In order to reduce the storage space and improve the data availability, some distributed storage systems adopt error correcting code technology to improve data availability and reduce the data storage space. Through analysis of MooseFS distributed file system, this paper proposes a method of error correction code implementation based on MooseFS. Through the test of data storage efficiency, this method can guarantee the commonly used “hot data” stored with multiple copies, “cold data” stored with the error correcting code, which greatly reduces space footprint and ensures the reliability at the same time.

Key words: distributed file system, data fault-tolerance, MooseFS, error-correcting code