Research And Implement On Technologies In Cluster Middleware Of Spatial Database

The amount of spatial data is increasing rapidly with the growth of spatial information acquisition and processing technology, and spatial database systems in many areas have a wide range of applications. Face to massive spatial data management, the traditional stand-alone database is difficult to meet the needs of multi-user frequently visits, calculation of complex geographical and so on. The database cluster technology becomes a hot spot. The main objective of the database cluster system is to achieve high performance, high availability and high confidence. And it is widely used in many fields such as high performance computing, massive data storage and management, Web services, e-commerce and so on.Spatial database cluster connected by a series of network connections need to handle a large number of user requests. The technology of load balancing is an effective approach to improve the utilization of computing resources and system performance. It distributes between the back-end database client requests, in order to reduce system bottlenecks, and enhance response capability of the system. Effective spatial data storage and distribution strategy is the key technology to improve the system availability. In addition, spatial database cluster middleware also need to support spatial data storage, which include vector data, raster data and other formats.In general, the paper focuses on the technology of spatial database cluster middleware. Firstly, we analyze the state of art of spatial database cluster management. We compare two kinds of systems: the cluster-based database engine and the cluster- based middleware. On the aspect of load balancing, after analyze load balancing technologies in detail we design the system structure of database cluster middleware which support for spatial data, based on the open-source database cluster middleware. Furthermore, we design the load balancing system capable of dynamic load balancing and resource measurement, and we implement a new algorithm for dynamic load balancing. In this paper, we describe the design idea of load balancing system, load balancing measure and the algorithm of load balancing. Finally, we validate the feasibility and validity of the new dynamic load balancing method by experiment studies. For spatial data management, we implement the storage and distribution strategy for spatial data based on database cluster middleware. Taking into account the characteristics of spatial data and cluster middleware, we develop different storage methods from them, respectively. Finally, we present our solutions for cluster management and spatial data visualization. Besides, we implement the cluster management system and spatial data import tools.