云计算环境下虚拟资源自适应调度策略

Strategy for Virtual Resource Scheduling Algorism in Cloud Computing Environment

由于云计算环境下的资源调度与以往网格调度存在巨大差异,提出了一种适应云计算环境的虚拟资源调度方法;首先定义了虚拟资源调度数学模型,然后给出了一种改进的遗传算法,该算法采用经典网格任务调度算法Min-min获取初始最优解,通过海明距离约束产生初始种群,并将调度模型对应的目标函数改进为适应度函数,对交叉算子、变异算子、交叉概率和变异概率等都进行了改进;最后,通过实验证明文中方法能获得全局最早完成时间,与其它方法相比,文中方法所求解的最早完成时间提高了近20%,是一种适合云计算环境的虚拟资源调度方法。

Aiming at that the resource scheduling has the big difference with the traditional grid task scheduling, a virtual resource scheduling method suiting to cloud computing was introduced. Firstly, the adaptive scheduling mathematical model was defined. Then aim proved Gene algorism was given, the algorism using the Min min algorism to get the initial solution, using Hamming distance as. the con straint [or the initial population, and the goal function of the scheduling model had changed to the fitness function, the cross operator, muta tion operator, cross probability and mutation probability were changed accordingly. Finally, the experiment shows the method in our paper has the optimal earliest completion time, and compared with other methods 20% is advanced and it is a virtual resource scheduling method suitable for cloud computation.