Radmin: Early Detection of Application-Level Resource Exhaustion and Starvation Attacks

Elsabagh, Mohamed; Barbará, Daniel; Fleck, Dan; Stavrou, Angelos

doi:10.1007/978-3-319-26362-5_24

Cited by 5 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each bucket is a linked list of entries holding counts for distinct words that hash to the same bucket. As each word is scanned from the input, the program invokes the add_word function (Lines [22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40]. This function first computes a hash value for that word-implemented in compute_hash (Lines 14-20)-and then attempts to find an existing entry for that word (Lines [28][29][30][31][32][33][34][35][36][37].…”

Section: Overview 21 a Motivating Examplementioning

confidence: 99%

“…Crosby and Wallach were the first to demonstrate denial-ofservice (DoS) attacks that exploit algorithmic complexity vulnerabilities [23]. Subsequent work on detecting and preventing DoS attacks [9,25,64] has typically focused on measuring aggregate resource exhaustion and does not specifically identify input characteristics that exploit worst-case algorithmic complexity.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

PerfFuzz: automatically generating pathological inputs

Lemieux

Padhye

Sen

et al. 2018

Proceedings of the 27th ACM SIGSOFT International Symposium on Software Testing and Analysis

132

View full text Add to dashboard Cite

Performance problems in software can arise unexpectedly when programs are provided with inputs that exhibit worst-case behavior. A large body of work has focused on diagnosing such problems via statistical profiling techniques. But how does one find these inputs in the first place? We present PerfFuzz, a method to automatically generate inputs that exercise pathological behavior across program locations, without any domain knowledge. Perf-Fuzz generates inputs via feedback-directed mutational fuzzing. Unlike previous approaches that attempt to maximize only a scalar characteristic such as the total execution path length, PerfFuzz uses multi-dimensional feedback and independently maximizes execution counts for all program locations. This enables PerfFuzz to (1) find a variety of inputs that exercise distinct hot spots in a program and (2) generate inputs with higher total execution path length than previous approaches by escaping local maxima. PerfFuzz is also effective at generating inputs that demonstrate algorithmic complexity vulnerabilities. We implement PerfFuzz on top of AFL, a popular coverage-guided fuzzing tool, and evaluate PerfFuzz on four real-world C programs typically used in the fuzzing literature. We find that PerfFuzz outperforms prior work by generating inputs that exercise the most-hit program branch 5× to 69× times more, and result in 1.9× to 24.7× longer total execution paths. CCS CONCEPTS • Software and its engineering → Software testing and debugging; Software performance;

show abstract

Section: Overview 21 a Motivating Examplementioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

PerfFuzz: automatically generating pathological inputs

Lemieux

Padhye

Sen

et al. 2018

Proceedings of the 27th ACM SIGSOFT International Symposium on Software Testing and Analysis

132

View full text Add to dashboard Cite

show abstract

“…A second example concerns the security domain, where several distributed applications can suffer from the so-called Denial of Service (DoS) attacks. According to [36] and [21], a DoS attack causes starvation because it deprives the resources of a target victim to provide services to legitimate users. Research in this domain is typically conducted from an empirical point of view, however we believe that, due to their complexity, such issues can benefit from a formal approach.…”

Section: Techniquesmentioning

confidence: 99%

An ASM-based characterisation of starvation-free systems

Bianchi

Pizzutilo

Vessio

2017

International Journal of Parallel, Emergent and Distributed Sys

View full text Add to dashboard Cite

State Machines (ASMs) have been successfully applied for modeling critical and complex systems in a wide range of application domains. However, unlike other well-known formalisms, e.g. Petri nets, ASMs lack inherent, domain-independent characterizations of computationally important properties.Here, we provide an ASM-based characterization of the starvation-free property.The classic, informal notion of starvation, usually provided in literature, is analyzed and expressed as a necessary condition in terms of ASMs. Thus, we enrich the ASM framework with the notion of vulnerable rule as a practical tool for analyzing starvation issues in an operational fashion.Keywords: abstract state machines; formal verification; starvation IntroductionSeveral formalisms are successfully applied to the development of critical and complex systems in a wide range of application domains, and to their ex-ante and ex-post analysis aimed at verifying and validating functionality and quality issues. Representing the system-under-study at a high level of abstraction allows developers to focus on algorithmic aspects, rather than on specific realizations of solutions at lower levels.Moreover, the mathematical foundation of formal methods provides complete and unambiguous investigations about the behavior and the properties the system-understudy is required to exhibit.Some formalisms provide inherent characterizations of properties -in the sense that they can be viewed as independent from the application domain -so that the formal verification of the computationally interesting properties of the modeled systems can be easily conducted. For example, in the Petri net framework [16], a marking Mi is reachable from an initial marking M0 if a sequence of transitions transforms M0 into Mi.If a marking is not reachable, then the transitions it drives are useless and can be deleted. However, several other formalisms do not provide such features.Our long-term research is aimed at providing an analogous framework for capturing computationally interesting properties with Abstract State Machines (ASMs)[28]. The goal is to enrich the general body-of-knowledge of the ASM framework and reinforce it as a conceptual tool that developers can find useful and practical in order to analyze system properties in an operational fashion. In this paper, "operational" means that the formal specification describes procedurally the system behavior by providing an abstract machine, which can be transformed in an executable form: this is the case of ASM-based models. This notation is usually mentioned in opposition to "declarative"

show abstract