大规模云计算环境虚拟机关联的内存故障数据集

大规模云计算环境虚拟机关联的内存故障数据集，包含多种虚拟机编号和内存故障数据。其中，内存故障数据包括原始特征以及基于专业知识处理后的特征。数据集采集于约30万个控制节点，由**云计算公司提供。数据集开放于亚太最大数据集管理平台——天池，目的是利用大规模的故障数据集研究故障预测算法，希望实现先进算法与实际生产环境接轨、实现故障预测问题与机器学习融合，促进云计算领域稳定性方向的发展和技术交流。

A large-scale memory failure dataset associated with virtual machines (VMs) in cloud computing environments, which includes multiple VM identifiers and memory failure data. Specifically, the memory failure data contains both raw features and features processed based on professional domain knowledge. The dataset was collected from approximately 300,000 control nodes and provided by a cloud computing company. It is hosted on TianChi, the largest dataset management platform in the Asia-Pacific region. The goal of releasing this dataset is to support research on failure prediction algorithms using large-scale failure datasets, align advanced algorithms with real-world production environments, integrate failure prediction problems with machine learning, and promote the development and technical exchanges in the field of cloud computing stability.

搜集汇总

背景与挑战

背景概述

该数据集是一个关于大规模云计算环境虚拟机关联内存故障的脱敏数据集，包含约30万个控制节点的数据，时间跨度为9个月。数据集提供了训练集、宕机事件时间信息和灰度测试集，旨在通过机器学习算法预测虚拟机内存故障，提升云计算系统的稳定性。