Energy-oriented denial of service attacks: an emerging menace for large cloud infrastructures

Abstract

This work analyzes a new and very subtle kind of security threat that can affect large scale cloudbased IT service infrastructures, by exploiting the computational resources of their component data center in order to waste as much energy as possible. The consequence of these threats range from increased costs in the energy bill, to penalization for exceeding the agreed quantity of green house gases (GHG) emissions, up to complete denial of service caused by electrical outages due to power budget exhaustion.We analyzed the different types of such attacks with their potential impacts on the energy consumption, modeled their behavior and quantified how current energyproportional technologies may provide attackers with great opportunities for raising the target facility emissions and costs. These efforts resulted in a simple model with some parametric reference values that can be used to estimate the impact of such attacks also in presence of very large infrastructures containing thousands or

A Belief Rule Based Expert System for Datacenter PUE Prediction under Uncertainty

“…However, this paper aims to extend this work by incorporating the use of a belief ruled based expert system for the prediction of a data center PUE with uncertainty. It ought to be noted that energy-related security in the context of a data centre [4] is not the focus of this study, but our future work might cover those aspects at a later stage. This remainder of this article is organized according in the following way: Section 2 covers a literature review, while Section 3 contains an overview of Belief Rule Based Expert Systems.…”

A survey and taxonomy of DoS attacks in cloud computing

“…The malicious activities of this attack cause a high workload on the target and keep them fully busy. The results of this attack are the increased costs of the energy consumption and penalization because of the greenhouse gas emissions for the cloud providers .…”

Evaluating the impact of eDoS attacks to cloud facilities

“…This is obtained in small scale by preventing single components from entering low consumption states, and in big scale by forcing the infrastructure to use more computing nodes than actually needed [4]; moreover, the storage subsystem could also be involved, with more additional energy requirements, and potentially additional network related costs, as in many architectures computing and storage nodes are separated and more network traffic implies more routing activities. As second, all components are overused: consequently, maintenance operations will be more frequent and parts will need to be replaced earlier than expected [12]. During a eDoS, the attacker bases its strategy on the request of services, that cannot be distinguished from other requests, with the double aim of monitoring the response of the system and slightly increasing the resources that are needed to serve them.…”

“…Eventually, a well designed eDoS attack can exploit the elasticity mechanisms of the cloud infrastructure, causing additional damage. The first examples of similar attacks are presented in [7] and [18], while a good survey on the topic is [12] and [6] presents a proof of concept and an experimental performance evaluation. Detecting an eDoS is generally difficult, as there is need for knowledge about temporal behaviour of energy usage and for good hypotheses about the regularity of the workload: while for batch systems or scientific oriented computation facilities a fairly good knowledge of the workload is conceptually achievable, the characterization of the workload of a typical cloud infrastructure is less easy to be available.…”