Large-scale and complex cloud computing systems are susceptible to software and hardware failures, which sig- nificantly affect the cloud performance and management. It is imperative to understand the failure behavior in cloud computing infrastructures. In this work, we study the impact of virtualization, which has become an enabling technology for cloud computing, on the cloud dependability. We present a cloud dependability analysis (CDA) framework with mech- anisms to characterize failure behavior in virtualized envi- ronments. We exploit failure-metric DAGs to analyze the correlation of various cloud performance metrics with failure events in virtualized and non-virtualized systems. We study multiple types of failures, including CPU-, memory-, disk- , and network-related failures. By comparing the generated DAGs in the two environments, we gain insight into the effects of virtualization on the cloud dependability. We also use the identified metrics to detect failures. Experimental results show that our approach can achieve high detection accuracy by using a small number of metrics. The proposed cloud dependability analysis framework is an open framework. Many analytical methods can be explored to implement the framework. We study the failure- metric DAGs, because of its competence in capturing the correlation of cloud performance metrics with failure events. Other methods might generate even better results. We plan to evaluate their performance on analyzing the cloud depend- ability and search for the best one(s) for future implementa- tion of the CDA framework. The scale of the current cloud computing testbed is relatively small compared with that of