NimbusGuard: A Novel Framework for Proactive Kubernetes Autoscaling Using Deep Q-Networks
Chamath Wanigasooriya, Indrajith Ekanayake · 2026
Read NimbusGuard: A Novel Framework for Proactive Kubernetes Autoscaling Using Deep Q-Networks by Chamath Wanigasooriya, Indrajith Ekanayake (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-11017.txt.
Description
Wanigasooriya and Ekanayake present NimbusGuard, an open-source proactive autoscaling system for Kubernetes built on deep reinforcement learning. Standard autoscalers react only after demand is detected, risking costly over-provisioning or slow under-provisioning. NimbusGuard forecasts workload with a long short-term memory model feeding a deep Q-network agent, scaling ahead of demand. Tests against the built-in Horizontal Pod Autoscaler and the event-driven KEDA autoscaler show the proactive approach performs better, cutting waste while protecting responsiveness.
Catalog record
| CREATOR | Chamath Wanigasooriya, Indrajith Ekanayake |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23764 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.11017