With the advancement of Internet technology, people's requirements for data computing and processing are getting higher and higher, and it is difficult for a single computer to meet efficient task computing and data transmission. Therefore, multiple computers are connected to complete a large-scale project. The needs of the task are becoming more and more urgent. With the increase of this demand, the distributed system(DS) emerges as the times require, and the mobile agent technology provides a new way to solve the distributed computing(DC) problem. For the system proposed in this paper, the service quality of computing tasks is evaluated by the transparent memory(TM) usage effect of the DC system. The experiments show that the TM algorithm is the most robust to the amount of memory contributed by nodes; and then through the multi-queue node scheduling experiment, it is found that the low priority The response time of the task is higher than that of the high priority task.