DLIDX: A Double Level Index for Cloud Storage Environment

摘要

As data grows in amount, the researchers focus on the multiple condition query problem in cloud store data. The traditional index structures which are used in the relationship database. They have defects in maintenance and query efficiency in the cloud storage environment. This paper proposes a double level index framework(DLIDX). In DLIDX, the index of cloud storage is constructed by global index and local indexes. Each property of cloud data forms a dimension. For each dimension, the global index establishes the mapping relationship between the property and the node devices by inverting the hash table. The local indexes, a kind of bitmap index, are stored in the node devices. The bitmap indexes support local query of nodes. This framework can reduce the unnecessary retrieval cost. Meanwhile, the executive efficiency of the framework is less affected by the dataset. The experimental result proves that the performance of DLIDX is better than other methods.