Design principles and features of analytic Space Data Grid

In terms of architecture, data grids have the following features: heterogeneity, scalability, self-applicability, and multi-GIS-level management domains. Based on the above features and actual application situations, when designing a data grid, follow these principles: String 4

(1) It has nothing to do with underlying implementation mechanisms, including data storage methods and metadata access methods. The system can define some interfaces or interfaces to encapsulate Different The storage system is heterogeneous in terms of data storage, directory management, and data transmission algorithms. String 6

(2) unrelated to application policies, the implementation of some important performance-related GPS functions should be left to users or applications for selection, rather than system encapsulation and execution, this not only ensures the flexibility of the user's implementation policies, but also improves the performance of the application system. For example, data moving and copying file directory management is a basic function of the Data Grid, the replication policy established on it should be determined by the user or application. String 9

(3) It is compatible with the grid framework. As an infrastructure for interconnection of wide-area heterogeneous resources, it provides many basic services for upper-layer systems, such as authentication, resource management, and information services. The design of the data grid should be based on the grid framework and be compatible with the underlying grid mechanism. String 1

(4) consistent information structure: In the grid, users access the structure information and status information of resources in a consistent manner. Similarly, in the data grid, users must also be able to access various information in the Data Grid in a consistent manner, such as metadata and copy file directories.