Abstract: Most cloud data centers use static resource scheduler algorithm to allocate and schedule virtualized resources. Once the scheduling is completed, if there is no maintenance, the resource allocation would not change. However, with the continuous use of resource pools, the problem of unbalance resource use becomes more and more serious. To solve this problem, this paper proposed a load balancing solution based on the Dynamic Resource Scheduler (DRS) algorithm to implement a relatively balanced resource usage on the physical nodes of the data center. The algorithm was used to monitor the resources in real time, select the physical hosts with large deviation from the average load value according to the monitoring results, calculate the final score combined with the resources allocated and actually used by the virtual machine, and select the appropriate virtual machine for dynamic resource balancing according to the score, so that the load value of the physical host tends to the average load value. The test results show that the DRS algorithm is effective and stable in the application of data center. It can dynamically schedule resources periodically, and finally achieve the goal of relatively balanced load in the data center.