Modeling the Impact of Intentional Behavior in Supercomputing Environments via 1-D and 2-D Stochastic Fluid Queues Fed by an ON-OFF Source

Stochastic Fluid Queues;Reliability Analysis;Supercomputing Environments;Applied Mathematics & Statistics

Liu, NingClark, Antwan ;

Johns Hopkins University

关键词: Stochastic Fluid Queues;    Reliability Analysis;    Supercomputing Environments;    Applied Mathematics & Statistics;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/59218/NLiu-thesis-master.zip?sequence=4&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

High performance computers (HPCs) are being used within a wide range of academic, commercial, and governmental applications. However, the resilience of these systems is in question making them especially vulnerable to various forms of questionable activity. Our work explores modeling these intentional actions as a first-order stochastic fluid queue (SFQ) fed by an ON-OFF source. Here, the constant flow rates depict failures, resulting from the intentional actions, and the source represents the failure injections as a result of the users;; actions. By performing spectral analysis techniques, the key contributions of this work are two-fold. First, closed-form solutions are derived depicting the probabilistic actions. Second, the transient analysis illustrates that the stability of the behavior depends on the combination of the flow rates, the state transition rates, and the Fourier modes. Numerical examples are provided which show that the proposed approach offers more granularity than prior works in SFQs that employ transform techniques. The results of this work can be leveraged to develop advanced frameworks for detection, analysis, and monitoring of Distributed Denial of Service (DDoS) attacks within supercomputing environments.