Load Balance Mechanism for Cloud Resource Provision Based on Virtual Machine Migration

摘要

The diversity of application demands and the heterogeneity of resources may lead to load imbalance in cloud resource provision. Aiming at this problem, a Virtual Machine (VM) resource provision genetic algorithm based on load balance, VMPGALB is presented in this paper. VMPGALB adopts a tree coding scheme, which is more suitable to reflect the characteristics of VM resource provision. When design the selection strategy, the proportion selection strategy based on fitness and the elite-preserving strategy are applied, which make individuals with smaller fitness be selected possibly and retain the best individuals directly to the offspring. For devising the crossover operator, VMPGALB has a better crossover performance through the crossover operation of two parent individuals and applying the spanning tree method. Meanwhile, in order to avoid a local optimum, VMPGALB enforces mutation operations on individuals according to a certain percentage. The experimental results show that VMPGALB could not only get better genetic performance and fewer VM migration numbers, but also gain the load balance scheme of VM resource provision with a faster convergence speed.