基于大数据的云数据中心智能运维系统

Intelligent Operation and Maintenance System of Cloud Data Center Based on Big Data

针对云计算中心容量趋势无法预测和异常无法监测的问题,基于ARIMA模型的时间序列算法和Isolation Forest的异常检测算法,提出一种以大数据分析为基础对云计算中心进行智能运维的方法。实验表明,基于Isolation Forest的服务器异常检测方法准确率约为88%,基于ARIMA模型的磁盘容量预测平均绝对误差为0.1586,证明了该系统相较于传统运维平台具有更好的稳定性、健壮性及服务能力。

Aiming at the problem of unpredictable capacity trends and inability to monitor anomalies in cloud computing centers,a method for intelligent operation and maintenance of cloud computing centers based on big data analysis is proposed using ARIMA time series algorithm and Isolation Forest anomaly detection algorithm.The experiment shows that the accuracy of the server anomaly detection method based on Iso⁃lation Forest is about 88%,and the average absolute error of disk capacity prediction based on ARIMA model is 0.1586,proving that this sys⁃tem has better stability,robustness,and service capabilities compared to traditional operation and maintenance platforms.