Fairness-aware Green Resource Scheduling Research In Cloud Environment

Since the concept of cloud computing was proposed, due to the driven by the marketing demand, the cloud service market has expanded rapidly, users scale has grown rapidly, making the need to meet the needs of users of massive resources,which led to the energy consumption of datacenter has become a widespread concern,cloud resource scheduling is the effective technology to solve this problem, but in the heterogeneous cloud environment, users are generally "greedy", some computing nodes in order to complete the calculation quickly, it will occupy other computing resources in a long time, destruct the fairness of resource scheduling strategies.Based on the analysis of the domestic and foreign scholars on the green cloud computing resource scheduling strategy and research on the present situation of fair distribution strategy, in view of the above problems, this dissertation has done the following work:(1) Due to the different types of resources in the cloud datacenter, the tasks are classified according to the proportion of the different types of its resource requirements, and distributed to the corresponding physical machine nodes to improve the utilization of the virtual machine resources.(2) The fairness calculation model is established to calculate the fairness value of the current system of the cloud computing datacenter in real time. The crossover operation of the standard genetic algorithm and the chromosome selection strategy of the next iteration are optimized to improve the resource utilization rate of the system.The fairness-aware multi-resource type allocation algorithm is proposed to ensure the utilization rate of various resources of the physical server is reasonable, it also verifies the feasibility and effectiveness of the algorithm.(3) This dissertation analyzes the current energy consumption of the cloud computing datacenter, it establishes the energy saving model which is suitable for the energy consumption of the cloud computing center for the physical server energy consumption and the virtual machine energy consumption. Under the condition of ensuring the fairness of the multi-type resource of the physical machine, the virtual machine migration of the physical machine is switched between the three states of sleep, shutdown and running, and an energy-efficient and fairness-aware resource scheduling algorithm is proposed, which can save energy expenditure while ensuring the fairness of cloud computing data center.(4) The existing scheduling model is simulated by open source CloudSim cloud simulator, compared the fairness value of each time point and the system energy consumption value of the cloud computing datacenter in the case of increasing the size of the virtual machine resources with the greedy algorithm, the Round Robin scheduling algorithm and the standard genetic algorithm in solving the double target resource to verify the superior performance of the scheduling strategy.