ZEST-Cloud: Zero-Trust, Energy-Aware, SLA-Sensitive Placement for Multi-Cloud Kubernetes
Gyanshauryam International Scientific Refereed Research Journal · 2023
DOI: 10.32628/gisrrj236433 ↗Kurz gefasst
Fachartikel in der Zeitschrift Gyanshauryam International Scientific Refereed Research Journal, 2023. Verfasst von Prudvi Saisaran Ponduru Prudvi Saisaran Ponduru und 1 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben.
Abstract (Originalsprache)
Cloud-native applications are increasingly placed across regions and providers to improve resilience, latency, and operational flexibility, but placement decisions also create security, cost, and environmental trade-offs. This paper introduces ZEST-Cloud, a security-first placement policy for multi-cloud Kubernetes. ZEST-Cloud treats zero-trust identity, mTLS, auditability, and data-residency requirements as hard eligibility constraints; only compliant regions enter the optimization stage. The remaining candidates are ranked using normalized response latency, modeled operational carbon, infrastructure cost, regional risk, projected load, and a quadratic service-level-objective penalty. The design follows the identity-centered direction of NIST SP 800-207 and the multi-cloud cloud-native guidance in NIST SP 800-207A. A fully executable discrete-time simulation was run with ten locked random seeds and 4,000 workloads per seed for the base experiment, plus burst-load, carbon-forecast-noise, and tighter-policy stress tests. In the base experiment, ZEST-Cloud achieved 67.28±0.36 ms mean latency, 114.49±1.07 ms p95 latency, a 2.48±0.32% SLA-violation rate, 12.71±0.18 kg of modeled operational CO2, $165.55±$1.99 modeled cost, and zero security-policy violations. Relative to latency-first placement, it reduced modeled carbon by 7.6% and cost by 1.3% while adding 9.95 ms mean latency. Relative to cost-first placement, it reduced mean latency by 38.0% and SLA violations by 34.5 percentage points, at 10.3% higher cost. A balanced optimizer without the hard trust gate appeared competitive on several operational metrics but violated security policy for 17.84% of requests. Ablations show that removing carbon awareness increases modeled emissions by 10.6%, while removing the explicit SLO penalty raises violation frequency. The results support a design principle rather than a universal optimum: mandatory trust rules should be enforced before, not traded against, cost, latency, or carbon objectives. All numerical values in this manuscript come from the accompanying executable simulation; no production-cloud measurement is claimed.
Zitieren
APA
Ponduru, P. S. P. P. S., & Nandanavanam, P. P. V. N. P. P. V. (2023). ZEST-Cloud: Zero-Trust, Energy-Aware, SLA-Sensitive Placement for Multi-Cloud Kubernetes. Gyanshauryam International Scientific Refereed Research Journal. https://doi.org/10.32628/gisrrj236433
Vancouver
Ponduru PSPPS, Nandanavanam PPVNPPV. ZEST-Cloud: Zero-Trust, Energy-Aware, SLA-Sensitive Placement for Multi-Cloud Kubernetes. Gyanshauryam International Scientific Refereed Research Journal. 2023. doi:10.32628/gisrrj236433
BibTeX
@article{Ponduru2023,
author = {Ponduru, Prudvi Saisaran Ponduru Prudvi Saisaran and Nandanavanam, Pavani Priya Vyshnavi Nandanavanam Pavani Priya Vyshnavi},
title = {ZEST-Cloud: Zero-Trust, Energy-Aware, SLA-Sensitive Placement for Multi-Cloud Kubernetes},
journal = {Gyanshauryam International Scientific Refereed Research Journal},
year = {2023},
doi = {10.32628/gisrrj236433}
}
EUVIMED