Abstract: Based on the study of cloud computing and mobile search engine, according to the development of current technology, we put forward the cloud retrieval system model based on clusters. The model consists of the cloud information layer, cloud retrieval cluster system, the user ctuery box We deeply studied distributed data storage technology in cloud retrieval system, and proposed program of cloud storage and retrieval of data structure. In order to improve the efficiency of parallel execution in cloud retrieval, put forward a program flow at the core of cloud retrieval software implementation of the model. The testing results of the model arc shown that the system function can realize correctly, and its performance is good and steady. The demonstration of this model gives the retrieval technology of massive information a scheme to develop thinking.

Key words: Cloud computing, Cloud retrieval, Program flow, Distributed file system, Cluster