Partition-based cloud data storage and processing model

Abstract

How to maximally use and efficiently manage cloud storage resources has become one of major problem in cloud computing area. If cloud storage capacity is limited or restrictions on a single file size, user may hire more cloud computing storage services. Data partitioning and distributed storage technologies are focused on to solve the problem that size of large dataset is limited in cloud storage. A model of partition-based data storage in cloud is put forward and its availability is discussed. Fragment partition description is given and three algorithms are designed on the basis of model. The first algorithm is about partition and storage of data, the second algorithm is about data query and connection, and the last algorithm is about fragment updated and re-partitioned. The experimental results show that the model can be applied to the situation with multiple cloud storage services and large dataset or file.