Given the complex dynamic and distributed nature of the cloud environment together with the discrete characteristic of cloud data center entities, we propose a dynamic scheduling strategy of cloud resources based on the distributed discrete event systems (DES) theory. First, we construct the cloud environment as a distributed DES composed of numerous subsystems, which include the physical server subsystem, the virtual machine subsystem, and the cloud task subsystem. Then, we apply supervisory control theory of distributed DESs to deal with the dynamic scheduling of the cloud resources. Through the control from local controllers and the information exchange between the global and local controllers, the proposed strategy achieves the overall optimization of the cloud resources and guarantees the rational allocation of the subsystems' resources. Moreover, the scheduling of the resources' load balancing improves the overall utilization of the cloud resources and the quality of the cloud service and enhances the performance of the entire cloud platform. Experiments indicate that with the increase in tasks, the proposed control strategy has obvious advantages in the utilization rate and the execution time compared with the traditional strategies.