Threat Modeling in Large-Scale Distributed Systems
Sai Prasad Veluru · International Journal of Emerging Research in Engineering and Technology · 2020
Since modern digital infrastructure is based on their broad distributed systems which include cloud platforms, global applications & also connected services the requirement of thorough threat modeling has become even more crucial. These systems provide a broad variety of possible vulnerabilities because of their size, heterogeneity & more dynamic properties even if they provide scalability, flexibility & also more resilience. The risk terrain is always changing and consists of improperly set up APIs, unsecured channels of their communication, insider threats, and sophisticated ongoing attacks. Prior to their spread into actual breaches, threat modeling provides a methodical technique for finding, evaluating & lowering risks. Still, huge scale implementation of this raises unique problems like the management of distributed trust borders, the preservation of visibility across components & the guarantee of consistency in security protocols. To fit the distributed paradigm, data flow diagrams, attack trees, STRIDE and DREAD models, and any other approaches have been changed or invented. This article investigates many approaches with an emphasis on their benefits & also disadvantages in pragmatic uses. We provide a case study of a global microservices architecture used by a financial organization to help to frame the discussion. The case study shows that early discovery of more vulnerabilities including privilege escalation channels and unsecured data propagation led to significant improvements to the general security posture of the system by means of their iterative threat modeling & mitigating strategies. The results draw attention to the need of multidisciplinary collaboration, automation in risk detection & also continuous review as systems grow. This work finally supports the integration of threat modeling as a continuous, basic activity in the design and administration of their distributed systems, therefore proposing a culture shift towards proactive security thinking instead of seeing it as a simple checklist