Limiting cache-based side-channel in multi-tenant cloud using dynamic page coloring

Abstract

Multi-tenant cloud, which features utility-like computing resources to tenants in a "pay-as-you-go" style, has been commercially popular for years. As one of the sole purposes of such a cloud is maximizing resource usages to increase its revenue, it usually uses virtualization to consolidate VMs from different and even mutually-malicious tenants atop a powerful physical machine. This, however, also enables a malicious tenant to steal securitycritical information such as crypto keys from victims, due to the shared physical resources such as caches.In this paper, we show that stealing crypto keys in a virtualized cloud may be a real threat by evaluating a cache-based side-channel attack against an encryption process. To mitigate such attacks while not notably degrading performance, we propose an approach that leverages dynamic cache coloring: when an application is doing security-sensitive operations, the VMM is notified to swap the associated data to a safe and isolated cache line. This approach may eliminate cache-based side-channel for security-critical operations, yet ensure efficient resource sharing during normal operations. We demonstrate the applicability by illustrating a preliminary implementation based on Xen and its performance overhead.

Virtual machine allocation policies against co-resident attacks in cloud computing

“…One way to encounter this kind of threat is to fundamentally eliminate the side channels between VMs, which is the approach taken by most previous work [9][10][11][12][13][14]. However, the proposed methods require substantial changes to be made to existing commercial platforms, and hence are impractical and not suitable for immediate deployment.…”

Exploring Virtual Machine Covert Channel via I/O Performance Interference

“…Later, more and more attack and defense work are conducted by researchers in this aspect. Besides the existed hardware based methods [24,25,26] to defend the CPU cache based timing channel, researchers proposed some novel approaches [11,13,12] to address the issue in the cloud environment; In the contrary, Xu et al [27] proposed the enhanced L2 Cache Covert Channels attacks that can improve the channel bit rate, they also show the limitation of CPU cache based timing channel. Independent but at the same time, Wu et al [3] proposed a practical covert channel method by exploiting the memory bus.…”

“…For example, malicious tenants can explore the CPU cache based side channel for private key extraction [2,4]. Soon after that, several approaches are provided to address the CPU cache based timing channel on virtualization platforms such as CPU cache usage detection and measurement [11], CPU cache partition [12] by dynamic page coloring, CPU cache non-sharing via system level protection [13] and etc. Meanwhile, Wu [3] et al also explores a practical memory bus based covert timing channel among resource sharing VMs.…”

Securing the Cloud: Threats, Attacks and Mitigation Techniques

“…More cloud-computing oriented research on the mitigation of side-channel attacks was introduced in [38]. This paper proposed an approach that leverages dynamic cache coloring.…”