Abstract: As the age of cloud computing,giant Web-scale application and massive data exert big challenges on centralined data retrieval. It becomes a tricky issue to process efficiently data retrieval in distributed environment. Traditional relational database is not fully suitable to the cloud any more. As a novel cloud storage, NoSQL (Not only SQL) emerges, among which Cassandra is widely used. With the application of index building under the distributed multi-node architecture, a scalable distributed inverted index built on Cassandra was put forward to solve the problem. Besides, analysis on data model and processing flow, as well as a performance test about Cassandra, were proposed.

Key words: Cloud storage, Distributed index, Inverted index, Cassandra