@inproceedings{66c1d557bc9447eaa4d6f2d3771502bf,
title = "Failure Prediction with Hierarchical Approach in Private Cloud",
abstract = "Cloud computing is widely adopted in real-world data centers. Most companies choose to build a private cloud service with the consideration of privacy. In these circumstances, they provide the service through Infrastructure as a Service (IaaS). However, with the scale of the data center, the possibility of cloud failure is increasing and become urgent in cloud computing. Current methods mainly use the proactive approach that monitors the failure and process it afterward. These methods are inefficient, and may always cause the service to break down. In this paper, we propose a new approach HFP (Hierarchical Failure Prediction) that can effectively monitor and predict the failure in advance. We firstly design and implement a new monitor structure that can effectively collect data. Then, a failure prediction method is proposed to predict the failure in advance. We implement this system with OpenStack, the synthetic result on the collected dataset shows that our method can achieve 98.3\% accuracy in the prediction of cloud failure.",
keywords = "Cloud computing, Failure prediction, Machine learning",
author = "Yaru Bao and Feilong Tang and Lijun Cao",
note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 15th International Conference on Green, Pervasive, and Cloud Computing, GPC 2020 ; Conference date: 13-11-2020 Through 15-11-2020",
year = "2020",
doi = "10.1007/978-3-030-64243-3\_35",
language = "英语",
isbn = "9783030642426",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "469--480",
editor = "Zhiwen Yu and Christian Becker and Guoliang Xing",
booktitle = "Green, Pervasive, and Cloud Computing - 15th International Conference, GPC 2020, Proceedings",
address = "德国",
}