The existing distributed database is mainly based on the limited scale network environment. Many systems implement algorithms are difficult to apply when it needs to be expanded or involved into the field of distributed query processing, distributed transaction processing and data replication. In this paper, one real-time transaction management architecture and its details are proposed to improve the efficacy of SQL processing based on distributed database system. The seven primary functional models in the server cluster part and four functional models in the distributed database part are designed to enhance the query function such as insert or update in our network environment. The experiments and results proved that our method is more efficient than traditional methods.