【摘要】 海量存储系统都采用元数据服务器机群的方式来处理文件系统的元数据信息。很多存储系统采用Hash算法来实现文件元数据在元数据机群内的分布,但是这些算法都是针对文件进行Hash。本文提出了一种目录哈希的新算法,针对目录进行Hash,并把一个目录内的元数据集中存储。本算法克服了文件Hash的不足,改善了存储系统的性能,并极大地提高了存储系统的可扩展性。更多还原

【Abstract】 Most large scale mass storage systems exploit metadata server clusters to handle the metadata of file system. Many storage systems use Hash algorithms to distribute the metadata among metadata server cluster. But the Hash key value of these algorithms is files. This paper proposes a new directory Hash algorithm. It treats directory as Hash key value, and implements a concentrated storage of metadata. This algorithm overcomes the disadvantages of file Hash algorithm. It can improve the performance of the storage system, and enhances the system’s scalability evidently.更多还原