Security Monitoring and Analytics in the Context of HPC Processing Model

Abstract

In this paper an overview of the problem of cybersecurity monitoring and analytics in HPC centers is performed from two intersecting points of view: challenges of assuring the necessary security level of HPC infrastructures themselves as well as new, not available earlier, opportunities to effectively analyze large volumes of heterogeneous data, facilitated by using large HPC clusters together with scalable analytic software. A major part of this paper is devoted to the most relevant methodologies and solutions that can be used by security analytics in order to at least partially face the challenge of analyzing large volumes of data potentially related with cyber-security events, in real-time or quasi-real-time. Particular solutions are considered in the context of their applicability in an HPC infrastructure. Relying on the results of experiments conducted within the SECOR project we have shown an approach of further development of the prepared architecture in HPC environment – within the confines of another R&D project, PROTECTIVE.