摘要
为降低大数据云中心的能量消耗和实现资源的优化配置,提出一种虚拟机资源高效分配策略;提出的策略对选定的特征上具备相似性任务分组的聚类进行定义,将各组任务映射到定制化的高效虚拟机类型;其高效指的是以最低限度的资源损耗成功执行任务;虚拟机的相关参数为核数量、内存量和存储量;虚拟机分配基于日志中提取的历史数据,并以任务的使用模式为基础;提出的资源分配策略以任务的实际资源使用量为基础,实现了能源消耗的降低;实验结果表明:不同聚类任务下,提出的虚拟机资源分配策略可以大幅节约能源消耗,具有较低的平均任务拒绝次数。
To reduce the energy consumption and optimize the allocation of resources in big data cloud center,a virtual machine resource allocation strategy is proposed.The proposed method defines the clustering of the selected features with similar task grouping,and maps the tasks of each group to the customized efficient virtual machine type.And this efficiency is the successful implementation of tasks with minimal resource depletion.The parameters of virtual machine are the number of cores,memory and storage.The virtual machine is based on the historical data extracted from the log trace,and it is based on the usage pattern of the task.The proposed resource allocation strategy is based on the actual resource usage of the task,and the energy consumption is reduced.The experimental results show that the proposed virtual machine resource allocation strategy can save energy consumption and reduce the average number of tasks in different cases of clustering.
出处
《计算机测量与控制》
2017年第8期272-274,314,共4页
Computer Measurement &Control
基金
湖南省科学技术厅科技计划项目(2011FJ3086)
关键词
大数据
资源分配
虚拟机
能量消耗
聚类
big data
resource allocation
virtual machine
energy consumption
clustering
