A Grading and Scheduling Method for Internal Data Flow in a Data Center

In order to achieve real-time and effective scheduling of the internal network traffic flow in data centers and ensure network service quality, this paper proposes a method for grading and regulating internal traffic flow in data centers. The traffic grading comprehensively considers the size and deadline of the flow, and calculates and updates the priority of the flow periodically, achieving a more accurate and reasonable division of traffic priority. At the same time, the optimal transmission path for the flow is calculated and selected based on two network congestion indicators: link utilization and queue length in the switch, avoiding the problem of uneven resource utilization in flow scheduling.