Deep-diving into an Easily-overlooked Threat: Inter-VM Attacks

‘Cloud computing’ has been embraced more than ever before due to a number of obvious advantages ( e.g. elasticity, space-saving, hardware consolidation and cost-saving) over traditional computing infrastructure. Virtualization is the core part of cloud computing as it provides a multi-tenancy model which is the foundation of cloud computing. Typically there are two types of threats inherent with virtualization environments. One wildly noticed threat is vm-escape which means a user on a virtual machine can ‘escape’ from it and take control over the whole hypervisor. The other threat known as inter-VM attacks have often been overlooked by the research community. In this paper we want to propose a framework which can systematically detect/analyze a number of inter-VM attacks. We want to build a system capturing such threats with a low false positive rate while providing valuable recommendations to system administrators.