EIPSIM: Modeling Secure IP Address Allocation at Cloud Scale


Public clouds provide impressive capability through resource sharing. However, recent works have shown that the reuse of IP addresses can allow adversaries to exploit the latent configurations left by previous tenants. In this work, we perform a comprehensive analysis of the effect of cloud IP address allocation on exploitation of latent configuration. We first develop a statistical model of cloud tenant behavior and latent configuration based on literature and deployed systems. Through these, we analyze IP allocation policies under existing and novel threat models. Our resulting framework, EIPSim, simulates our models in representative public cloud scenarios, evaluating adversarial objectives against pool policies. In response to our stronger proposed threat model, we also propose IP scan segmentation, an IP allocation policy that protects the IP pool against adversarial scanning even when an adversary is not limited by number of cloud tenants. Our evaluation shows that IP scan segmentation reduces latent configuration exploitability by 97.1% compared to policies proposed in literature and 99.8% compared to those currently deployed by cloud providers. Finally, we evaluate our statistical assumptions by analyzing real allocation and configuration data, showing that results generalize to deployed cloud workloads. In this way, we show that principled analysis of cloud IP address allocation can lead to substantial security gains for tenants and their users.

Kyle Domico
Kyle Domico
PhD Student and Research Assistant

My research lies at the intersection of reinforcement learning and computer security.