Research Of Multi-Root Tree And It's Applications In Distributed Environment

With the development of the distributed technology and the parallel computing, we found that the relational model was difficult to adapt to the distributed environment. Relational model is widely used in a centralized environment, but it is inadequate for the distributed environment. Multi-root tree was proposed in 1989 by the Chinese scholars. Many researchers also proposed the similar structure. A lot of researches have been applied in various fields and achieved some success. Multi-root tree is a significant hierarchical model and it can be a good description of the distributed environment.The main work includes:first, studies the relational model's schema graph and data graph, describes how to turn the relational model into a schema graph, and shows the definition of the data graph and the schema graph. Then we explores the nature of the multi-root tree, studies how to transform the schema graph and the data graph into multi-root tree and how to use multi-root tree to descript the distributed database. Also we describe conversion's methods and limits, study the data extraction operation of the relational model with the conditions that model related and unrelated, and further discusses the data extraction operation under a single control conditions and multiple conditions, data merge operation. Then further studies of the data extraction operation of relational model with the semantically related data, analysis different operating strategies, the extraction method of multi-root tree and the feasibility and advantages and disadvantages of these methods. We use the multi-root tree operation to optimize the operation of the relational model and add artificial intervention and other means.Finally, we combine with the project that waterway system, apply the data extraction method in the system, ensuring the data distribution and data consistency, and focus on validation the efficiency of the data extraction. We find the bottleneck and also give the corresponding optimization strategy. Then we show the broad application prospects of the multi-root tree.