Two-level data middle platform flow-batch integrated transmission strategy based on dynamic priority

State Grid two-level data center is a data management and service platform for smart grid, which can realize the functions of data collection, storage, processing, analysis and application. However, its data link transmission faces the challenges of high transmission delay, poor quality of service and difficult to transmit batch data cooperatively on the same data link. In order to solve these problems, this paper proposes an integrated flow and batch transmission configuration model. According to the characteristics and requirements of the data, the model comprehensively considers the factors such as data priority, real-time, urgency and data size, as well as the load and bandwidth requirements of the link, and realizes the data classification and priority division, weighted hybrid scheduling and link allocation transmission control. In order to verify the validity of the model, NS3 is used as the simulation platform, and simulation experiments are carried out in a network topology including headquarters data center, provincial data center and nodes of each business department. The simulation results show that under the same network conditions, compared with the traditional FIFO queue scheduling algorithm and WFQ queue scheduling algorithm, the model can increase the data link transmission efficiency by 20%, reduce the total delay by 21%, ensure the quality of service, and realize the coordinated transmission of stream and batch data.