Abstract:
As an important infrastructure of cloud computing, the problem of high energy consumption is becoming more and more serious. In order to reduce carbon emissions of data centers and achieve the goal of "carbon peak and carbon neutrality", data center operators begin to introduce new energy sources. However, the randomness and uncertainty of new energy output increase the difficulty of new energy utilization. Based on the spatio-temporal complementarity of new energy output in multiple geographically distributed data centers, a spatio-temporal scheduling algorithm for computing tasks in multiple data centers is proposed. The flexible scheduling of computing tasks can realize the migration of computing tasks in time and space, and then align the power curve of data center with the output of new energy, so as to realize the maximum utilization of new energy. In this paper, real data are used to analyze the algorithm, which verifies the effectiveness of the algorithm for new energy consumption and reduces the carbon emissions of the data center cluster.