TopoEvo: A Topology-Aware Self-Evolving Multi-Agent Framework for Root Cause Analysis in Microservices

2026-05-15

AI Summary

This work addresses the challenges of root cause analysis in microservice systems, where multimodal data noise, fault cascading-induced symptom amplification, and dynamic topology drift often lead existing methods to misidentify prominent downstream symptoms as root causes. To overcome these limitations, we propose TopoEvo, a novel framework that integrates topology-aware multimodal alignment (MOMA), symptom tokenization with propagation consistency verification, and a self-evolution mechanism driven by high-confidence pseudo-labels. This enables multi-agent collaborative hypothesis-evidence-validation reasoning in dynamic environments. By explicitly mitigating symptom amplification bias and adapting robustly to topological changes, TopoEvo achieves more accurate, resilient, and auditable root cause localization in complex microservice architectures.

Abstract

Root cause analysis (RCA) in microservices is challenging due to (i) noisy and heterogeneous multimodal observability (metrics, logs, traces), (ii) cascading failure propagation that amplifies downstream symptoms, and (iii) non-stationary topology drift induced by autoscaling and rolling updates. Recent LLM-based RCA agents can generate tool-grounded explanations, yet they often remain topology-agnostic and suffer from \emph{symptom-amplification bias}, misattributing the root cause to salient downstream victims. We propose \textbf{TopoEvo}, a topology-aware self-evolving multi-agent framework that couples graph representation learning with structured, topology-constrained reasoning. TopoEvo first introduces \emph{Metric-orthogonal Multimodal Alignment} (MOMA), which decomposes metric embeddings into complementary subspaces and contrastively aligns logs and traces to reduce modality redundancy and sparsity, yielding stable node representations for graph encoding. It then applies \emph{Vector Quantization} (VQ) to discretize topology-enhanced states into auditable \emph{symptom tokens} with a symptom lexicon, enabling reliable retrieval and token-level evidence grounding. On top of these discrete topology cues, TopoEvo performs a multi-agent \emph{Hypothesis--Evidence--Test} (HET) workflow to explicitly verify propagation-consistent explanations and separate initiating anomalies from amplified downstream symptoms. Finally, a \emph{Self-Evolving Mechanism} refreshes hierarchical incident memory and performs conservative test-time adaptation with high-confidence pseudo-labels to maintain robustness under drift.

Problem

• root cause analysis
• microservices
• topology drift
• cascading failures
• multimodal observability

Innovation

• topology-aware reasoning
• multimodal alignment
• vector quantization
• multi-agent RCA
• self-evolving framework