Research On Load Balancing In The Construction Of Cloud Computing Data Center

Electric power data center is the basic information platform of electric power company. Along with the continuous development of electric power company information system scale and the improvement of automation level, the disadvantages of traditional electric company data center such as low equipment utilization rate, limited availability, long restoration time when there is fault service, large loss of business has become increasingly prominent. City double live data center based on cloud computing is the future development trend of electric power data center.Due to the customer demands’diversity, dynamic change and the heterogeneity of the resource server, cloud computing data center appears unbalanced load conditions. This paper firstly introduces the architecture and key technology of city double live data center based on cloud computing. Define the main resources of cloud computing data center with the mathematical language. In order to solve the unbalanced resource allocation problem and low resource utilization rate problem caused by resources heterogeneity, this paper proposes a multidimensional virtual machine resource placement algorithm based on ant colony optimization algorithm. In order to verify the advantages of the proposed algorithm, we compare it with the virtual machine placement algorithm based on greedy strategy and basic ant colony in the open source cloud simulation platform CloudSim, which demonstrate the effectiveness of the virtual machine placement algorithm based on ant colony optimization algorithm. Then we propose a load balancing algorithm based on virtual machine dynamic migration. The algorithm uses the load prediction to trigger load adjustment, which avoid unnecessary virtual machine migration caused by instantaneous peak and reduces the system overhead. Then classify the physical machines according to the load data. Transfer the corresponding virtual machine on the high load physical machine to the low load physical machine. Adjust the number of data center physical machine according to the system load saturation. Verify the performance of the load adjustment algorithm by simulation experiment. Finally, analysis of the related needs of cloud resources scheduling simulation system and complete the module design and database design of the system platform. Develop the different components according to the design of the system. Finally complete the load balancing system and carries on the test and verification.