Abstract:

According to the characteristics of spatial data and their relationships, we proposed a load balanced spatial vector data layout target, and combined with the specific environment, using the algorithm of spatial data layout based on graph coloring theory and data security mechanism of multiple replica. Spatial data deployed in parallel environment, and improved the execution efficiency and data security of task on the data nodes. Experimental results show that the layout method can achieve balanced layout of data, take into account the efficiency and security of the parallel computing, and adapt to more query applications.

Key words: parallel computing system, spatial data layout, graph coloring theory, multi replica