FDSSS: An Efficient Metadata Management Scheme in Large Scale Data Environment

摘要

The performance of metadata query and search largely affects the storage system efficiency, especially in information lifecycle management (ILM) systems. In this paper we present FDSSS, a metadata management scheme, to optimize the metadata query and search performance, which stores key attributes of metadata in the main memory and organizes them as a tree-like structure. Through performance evaluation in large data scale (about 224 metadata items) environment, FDSSS improves the query and search performance than traditional metadata management which using database to store all the metadata items