A disruption in one region can lead to a disruption of the logistics network with severe economic impacts. In this study, we construct China's logistics network from the perspective of space of flows with the help of online freight information on the Yumman platform, and analyzes the dynamics of risk propagation in the logistics network through the cascade failure model. The results show that: (1) nodes with generally high centrality values of various kinds have the potential to trigger large-scale avalanches when they act as sources of risk propagation; (2) indirect edges in the logistics network play a significant role in the cascading failure process; (3) A load capacity higher than 1.4 times the volume of material flow is an effective means of preventing unexpected risks in Shanghai; (4) disruptions in the east and west hub cities will cause the logistics network in the western region to become paralyzed.