Hypothesize-Then-Verify: Speculative Root Cause Analysis for Microservices with Pathwise Parallelism

TL;DR

The paper tackles root cause analysis in complex microservice systems by introducing SpecRCA, a speculative RCA framework that uses a hypothesize-then-verify paradigm with pathwise parallelism to improve both accuracy and speed. It decomposes the process into Hypothesis Drafting, which generates diverse candidate causes from metrics, traces, and logs, and Root Cause Verifier, which validates each hypothesis in parallel using RCALite, augmented by Teacher LLM distillation and reinforcement learning. Preliminary results on the AIOps 2022 dataset show notable gains in accuracy (Recall@1: ~61%, Recall@3: ~76%, Recall@5: ~82%, MRR ~63%) and a substantial speedup (≈9.89s per query) over baselines. The work suggests a scalable, interpretable RCA workflow for real-time microservice environments and outlines concrete avenues for further acceleration and dataset expansion.

Abstract

Microservice systems have become the backbone of cloud-native enterprise applications due to their resource elasticity, loosely coupled architecture, and lightweight deployment. Yet, the intrinsic complexity and dynamic runtime interactions of such systems inevitably give rise to anomalies. Ensuring system reliability therefore hinges on effective root cause analysis (RCA), which entails not only localizing the source of anomalies but also characterizing the underlying failures in a timely and interpretable manner. Recent advances in intelligent RCA techniques, particularly those powered by large language models (LLMs), have demonstrated promising capabilities, as LLMs reduce reliance on handcrafted features while offering cross-platform adaptability, task generalization, and flexibility. However, existing LLM-based methods still suffer from two critical limitations: (a) limited exploration diversity, which undermines accuracy, and (b) heavy dependence on large-scale LLMs, which results in slow inference. To overcome these challenges, we propose SpecRCA, a speculative root cause analysis framework for microservices that adopts a \textit{hypothesize-then-verify} paradigm. SpecRCA first leverages a hypothesis drafting module to rapidly generate candidate root causes, and then employs a parallel root cause verifier to efficiently validate them. Preliminary experiments on the AIOps 2022 dataset demonstrate that SpecRCA achieves superior accuracy and efficiency compared to existing approaches, highlighting its potential as a practical solution for scalable and interpretable RCA in complex microservice environments.

Figures (3)

• Figure 1: The overview of SpecRCA. SpecRCA follows a Hypothesize-Then-Verify paradigm: it first employs a Hypothesis Drafting module to quickly generate potential hypotheses, and then uses a Root Cause Verifier to validate them in parallel.
• Figure 2: Hypothesis Drafting
• Figure 3: Root Cause Verifier