🤖 AI Summary
This work addresses the inefficiencies in cloud-native autoscaling caused by misalignment between business policies and resource scheduling, as well as the lack of coordination between pod and node scaling, which often leads to resource waste and performance degradation. To resolve these issues, the authors propose MAS-H2, a three-tier hierarchical multi-agent system that introduces, for the first time, a hierarchical multi-agent architecture to cloud-native autoscaling. MAS-H2 integrates strategic, planning, and execution agents to formalize business objectives, enable joint proactive scaling of pods and nodes, and achieve zero-downtime strategy migration. Implemented as a Kubernetes Operator, the system combines time-series forecasting, global utility optimization, and multi-agent control. Experimental results on GKE demonstrate that CPU utilization remains below 40%, sustained load is reduced by over 50% compared to HPA, peak CPU load during traffic bursts drops by 55%, and zero-downtime infrastructure migration is successfully accomplished.
📝 Abstract
Autoscaling in cloud-native platforms like Kubernetes is reactive and metric-driven, leading to a strategic void problem. This comes from the decoupling of higher-level business policies from lower-level resource provisioning. The strategic void, coupled with a fragmented coordination of pod and node scaling, can lead to significant resource waste and performance degradation under dynamic workloads. In this paper, we present MAS-H2, a new hierarchical multi-agent system that addresses the challenges of autonomic cloud resource management with a complete end-to-end solution. MAS-H2 systematically decomposes the control problem into three layers: a Strategic Agent that formalises business policies (e.g., cost vs. performance) into a global utility function; Planning Agents that produce a joint, proactive scaling plan for pods and nodes with time-series forecasting; and Execution Agents that execute the scaling plan. We built and tested a MAS-H2 prototype as a Kubernetes Operator on Google Kubernetes Engine (GKE) to benchmark it against the native Horizontal Pod Autoscaler (HPA) and Cluster Autoscaler (CA) baselines under two realistic, spiky, and stress-inducing workload scenarios. The results show that the MAS-H2 system maintained application CPU usage under 40% for predictable Heartbeat workloads. This resulted in over 50% less sustained CPU stress than the native HPA baseline, which typically operated above 80%. The MAS-H2 system demonstrated proactive planning in a volatile Chaotic Flash Sale scenario by filtering transient noise and deploying more replicas compared to HPA. It reduced peak CPU load by 55% without under-provisioning. Beyond performance, MAS-H2 seamlessly performed a zero-downtime strategic migration between two cost- and performance-optimised infrastructures.