ARGOS: Reinforcement Learning-Driven Multidimensional Elasticity for Service Orchestration in the Computing Continuum

2026-09-29

AI Summary

This study addresses the challenge of balancing analytics quality, cost, and resource consumption for data-intensive services operating under resource saturation across the computing continuum. To this end, we propose an end-to-end controller that formulates multi-dimensional elasticity as a Markov Decision Process. By leveraging deep reinforcement learning and heterogeneous cluster orchestration, the proposed method transcends conventional resource-only scaling to achieve, for the first time, dynamic adaptation of analytical requirements—such as coverage—and capacity-aware admission control. Experimental evaluations demonstrate that the proposed strategy outperforms non-learning baselines and closely approximates optimal fixed references. Furthermore, it significantly enhances service quality under saturated workloads while strictly avoiding CPU and memory violations.

Abstract

Data-intensive services in the Computing Continuum must balance analytics quality, resource usage, and cost across heterogeneous nodes with limited and uneven capacity. This balance becomes especially difficult when resource scaling reaches capacity limits, because changes in demand and cluster pressure must then be absorbed without violating client-defined quality ranges. Existing orchestrators mainly adapt resources, placements, or replicas, while analytics requirements such as coverage, sample, and freshness remain fixed. This article presents ARGOS, the Adaptive Reinforcement Learning-Driven Governance for Orchestrated Services, an end-to-end controller that formulates multidimensional elasticity as a per-request Markov decision process over analytics quality and cluster pressure, supported by capacity-aware admission. ARGOS is evaluated under controlled workloads and time-varying multi-tenant arrivals on a heterogeneous cluster. Across the controlled scenarios, the deep reinforcement learning policies consistently outperform the non-learning baselines and approach the independently tuned best-fixed reference. A separate live evaluation reports improvements over the static midpoint under realistic and saturated arrivals, with no recorded CPU or memory violations but remaining coverage violations. These results support deep reinforcement learning as an adaptive mechanism for multidimensional elasticity when resource scaling alone is insufficient.

Problem

• Computing Continuum
• Service Orchestration
• Multidimensional Elasticity
• Resource Scaling
• Heterogeneous Cluster

Innovation

• Deep Reinforcement Learning
• Multidimensional Elasticity
• Markov Decision Process
• Service Orchestration
• Computing Continuum