Russian Federation
The article presents the development and analysis of stochastic models of containerized application behavior in Kubernetes 1.29+ based on inhomogeneous Markov chains and hyperexponential tail latency distributions. The authors decompose stochasticity into processes, formalize pod transition matrices and implement vectorized simulation of millions of pods with online Prometheus calibration. The uniqueness of the work lies in p95-p99.9 tail latency prediction with <5% error, 15-30 sec lead autoscaling, hidden correlation detection and 40% placement density optimization without overloads.
stochastic containers, Kubernetes, tail latency, Markov chains, autoscaling
1. Mochalov, V. P. Algoritm dinamicheskogo raspredeleniya i balansirovki nagruzki v raspredelennyh oblachnyh vychisleniyah / V. P. Mochalov, N. Yu. Bratchenko, D. V. Gosteva // Modelirovanie sistem i processov. – 2024. – T. 17, № 1. – S. 92-102. – DOIhttps://doi.org/10.12737/2219-0767-2024-17-1-92-102. – EDN EWMPYM.
2. Haritonov, A. V. Priblizhennaya ocenka zaderzhek v komp'yuternoy sisteme s konteynernoy virtualizaciey // Informatika i avtomatika. – 2025. – T. 24, № 3. – S. 917-925. – URL: https://www.mathnet.ru/links/3bb517bfcc71c951793066d9968cc9cf/trspy1377.pdf (data obrascheniya: 19.02.2026).
3. Grigor'ev, V. P. Ocenka veroyatnostno-vremennyh harakteristik konteynerizirovannyh prilozheniy // NTV ITMO. – 2023. – № 2. – S. 249-258. – URL: https://ntv.ifmo.ru/file/article/22755.pdf (data obrascheniya: 19.02.2026).
