Hadoop资源信息智能监测仿真

Intelligent Monitoring Simulation for Hadoop Resource Information

针对当前Hadoop资源负载监测过程易出现能耗较高、时延明显的问题,提出高负载量移动服务器Hadoop资源信息智能监测方法。利用时间序列将Hadoop资源分化至多种资源访问序列中,分析序列类似度并估算类似值。利用计算最邻近节点方法划分高负载资源序列。计算高负载系数,获取特征向量集向量,将其转换为特征矩阵,传输高负载特征信息结果,完成高负载量移动服务器Hadoop资源信息智能监测。实验结果证明,研究方法能够有效对Hadoop资源信息进行高负载度监测,且监测过程能耗更低,响应延时问题得到明显改善。

At present,the monitoring process of Hadoop resources is susceptible to the problem of high energy consumption and obvious time delay.Therefore,an intelligent monitoring method for Hadoop resource information of high-load mobile servers was proposed.The time series was used to divide Hadoop resources into different resource access sequences.The similarity degree among sequences was analyzed and the similar value was estimated.The high-load resource sequence was divided by calculating the nearest node.The high-load coefficient was calculated to obtain the feature vector set,and then it was converted into the feature matrix to transmit the result of high load feature information.Thus,the intelligent monitoring of Hadoop resource information of the high-load mobile server was completed.Experimental results prove that the proposed method can effectively monitor Hadoop resource information with high load,and the monitoring process consumes less energy,so that the problem of response delay is significantly solved.