Abstract: With the development of cloud computing, the scale and complexity of the data center has risen to a new level. How to monitor the whole information system and get the real-time risk status of the system has become a hot research topic in both academia and industry. For this situation, this paper proposes a real-time risk assessment schema (RRAS) by acquiring real customers’ actual needs, combining risk assessment theory and using the analysis results of host logs. This paper simulates three types of attacks to verify RRAS. The results show that RRAS can assess the risk situation of entire system for real-time with high accuracy.

Key words: risk assessment, host log based, cloud data center, risk index, real-time

摘要： 随着云计算的发展，数据中心的规模和复杂性都上升到一个新的高度。如何对整个信息系统进行全面的监控，或者说如何实时准确地获知当前整个系统的风险状态已经成为很多企业和机构亟需解决的一个热点问题。针对这一问题，通过模拟真实用户的实际需求，结合风险评估的系统理论知识，利用主机日志的分析结果，提出了一种实时风险评估模型（real-time risk assessment schema，RRAS）。在实验环境中对RRAS进行了实现，并完成了3组模拟攻击实验。实验结果表明，该模型可以实时地对整个系统进行风险评估，并且得到的结果具有较高的准确性。

关键词: 风险评估, 基于主机日志, 云数据中心, 风险指数, 实时