Controlled chaos engineering experiments within cloud infrastructure demonstrate that virtual private cloud-based scaling mechanisms improve application resilience during intense HTTP denial-of-service attacks. Researchers conducted extensive controlled testing to evaluate how native architectural and scaling mechanisms sustain service dependability when cloud-native systems face aggressive traffic floods.
Cloud-native applications face mounting operational vulnerabilities from high-traffic HTTP flood attacks that can overwhelm automated scaling systems and severely degrade overall service availability. Conventional defence strategies predominantly focus on traffic filtering, detection, and reactive mitigation. However, these traditional models frequently overlook structural architectural approaches and resilience delivered through adaptive scalability mechanisms.
To investigate resilience by design, investigators deployed an empirical evaluation protocol that subjected cloud infrastructure to sustained attack conditions using chaos engineering principles. The experimental campaign was carried out within Amazon Web Services infrastructure, spanning more than 80 hours of controlled testing and generating over 10 million HTTP requests.
The experimental setup systematically compared four dedicated scaling and resilience mechanisms under attack: Default scaling, Warm-up scaling, 2-Desired capacity scaling, and virtual private cloud-based scaling. Researchers leveraged native cloud provider services to construct the test environments, incorporating auto-scaling groups, application load balancers, security groups, and multi-availability zone deployments.
Technical and statistical analyses revealed that while all tested scaling configurations behaved similarly under low-load conditions, their performance diverged significantly under high-intensity attacks. The virtual private cloud-based mechanism demonstrated superior resilience by reducing the request failure rate by 12 percent compared to the default scaling configuration.
The virtual private cloud-based mechanism also enhanced application availability while reducing observed total request volume by approximately 60 percent relative to default settings during HTTP floods. Furthermore, this architectural configuration successfully mitigated service disruption without generating excessive over-provisioning costs.
These experimental findings establish that scalability mechanisms serve as an effective resilience-by-design strategy that directly complements traditional security-based perimeter defences. By providing empirical guidance, the study enables cloud practitioners and software architects to prepare systems for high-volume floods while preventing unnecessary autoscaling expenses.
The investigators noted that future research should examine predictive auto-scaling models and real-time adaptive mechanisms to further strengthen cloud-native applications against evolving HTTP flood attacks.
Subject of Research: Cloud-native application resilience and scalability under HTTP flood attacks
Article Title: Empirical chaos-driven approach for cloud resilience and scalability under HTTP DDoS attacks
Article References: Al-Said Ahmad, A., Al-Qoraʼn, L. F., & Tildsley, H. (2026). Empirical chaos-driven approach for cloud resilience and scalability under HTTP DDoS attacks. Journal of Cloud Computing. https://doi.org/10.1186/s13677-026-00980-9
Image Credits: AI Generated
DOI: 10.1186/s13677-026-00980-9
Keywords: cloud computing, HTTP DDoS attacks, chaos engineering, cloud resilience, scalability mechanisms, Amazon Web Services, auto-scaling, application load balancers, virtual private cloud, service availability, failure rate reduction, traffic flood mitigation, multi-availability zone, cloud architecture, over-provisioning costs, predictive auto-scaling
News Source: Denise Maddox. (November 19, 2026). Chaos Engineering Experiments Reveal VPC-Based Scaling Mitigates Cloud HTTP Flood Attacks. Scienmag.



