LLM-Based Multi-Agent Collaboration for Constrained Multi-Objective Container Placement

2026-10-04

AI Summary

This study addresses the challenge of balancing power consumption and affinity in data center container placement, where conventional methods often exhibit poor adaptability and are prone to local optima. To overcome these limitations, this work proposes a large language model-based multi-agent collaborative framework. The framework integrates ReAct reasoning with candidate table filtering to generate feasible placement decisions, introduces Tolerance-Gated Arbitration (TGA) to resolve multi-objective conflicts, and employs a Monotonic Improvement Rule (MIR) to ensure strict satisfaction of hard constraints. Evaluated on the Google Cluster Trace dataset, the proposed approach achieves an 8.36% reduction in power consumption and a 39.30% improvement in affinity. Furthermore, comprehensive ablation studies validate the effectiveness of each individual component within the framework.

Abstract

Container placement in data centers must simultaneously minimize power consumption and maximize affinity preferences, while satisfying multi-resource capacity and anti-affinity constraints. Traditional approaches typically rely on fixed rules, which lack adaptability to dynamic cluster states and are difficult to extend for adaptive decision-making. On the other hand, meta-heuristic methods, although more flexible, are often computationally expensive, slower and prone to getting trapped in local optima. In this work, we propose an adaptive and efficient approach based on an LLM-driven multi-agent collaboration framework, where four specialized agents operate in a closed-loop ReAct cycle at each placement step. A Power Consumption Agent and an Affinity Agent debate over competing objectives, while a Placement Agent resolves conflicts using Tolerance-Gated Arbitration (TGA). A Rearrangement Agent further refines decisions through the Monotone Improvement Rule (MIR) in a post-placement refinement phase. All the agents reason over deterministic, feasibility-filtered candidate tables provided by the environment, ensuring that every proposed action inherently satisfies hard constraints. Based on experimental evaluation using the Google Cluster Trace with a configuration of 100 applications and 25 machines, the proposed framework consumes 8.36% less power and achieves 39.30% higher affinity than the power-greedy baseline. Ablation studies confirm that each architectural component, the multi-round debate mechanism, the TGA in the Placement Agent, and the MIR in the Rearrangement Agent, contributes meaningfully to the overall performance, with removal of any single component degrading both power consumption and affinity satisfaction.

Problem

• Container Placement
• Multi-Objective Optimization
• Constrained Optimization
• Data Center
• Power Consumption

Innovation

• Multi-Agent Collaboration
• Large Language Model
• Container Placement
• Tolerance-Gated Arbitration
• Monotone Improvement Rule