一句话：提出无监督框架 ART，借助共享知识统一完成微服务系统的异常检测、故障分诊与根因定位。
微服务系统的事件管理涵盖异常检测（AD）、故障分诊（FT）与根因定位（RCL），但现有技术多只针对单一任务建模，忽视任务间共享知识，导致效率低、响应慢、缺乏系统视角且更新运维复杂。本文提出无监督框架 ART，给出覆盖 AD、FT、RCL 的全流程统一方案。作者先通过实证研究分析异常偏差中的共享知识在三类任务中的体现方式；为度量偏差并抽取共享知识，依次用 Transformer Encoder、GRU、GraphSAGE 建模通道依赖、时间依赖与调用依赖；再构建带显式语义信息的统一故障表示，提升抽象特征的可解释性，并作为多任务无监督求解的基础。在两个基准微服务系统生成的数据集上，ART 相对现有方法在 AD、FT、RCL 上分别提升 5.65%–60.8%、13.2%–95.7% 与 13.3%–205%。
Automated incident management is critical for large-scale microservice systems, including tasks such as anomaly detection (AD), failure triage (FT), and root cause localization (RCL). Currently, most techniques focus only on a single task, overlooking shared knowledge across closely related tasks. However, employing isolated models for managing multiple tasks may result in inefficiencies, delayed responses, a lack of systemic perspective, and complexity in updates and operations. Therefore we propose ART, an unsupervised framework that integrates a full-process solution covering Anomaly detection, failure Triage, and Root cause localization. It reaches the unification of multiple tasks by extracting the shared knowledge. Specifically, we first conduct an empirical study to analyze how the shared knowledge embedded in anomalous deviations manifests in AD, FT, and RCL. To better calculate deviations and extract shared knowledge, we sequentially model channel, temporal, and call dependencies using Transformer Encoder, GRU, and GraphSAGE, respectively. Then unified failure representations enhance the interpretability of abstract features with explicit semantic information, serving as the basis for unsupervised multitask solutions. Our evaluations on the datasets generated from two benchmark microservice systems demonstrate that ART outperforms existing methods in terms of AD (improving by 5.65% to 60.8%), FT (improving by 13.2% to 95.7%), and RCL (improving by 13.3% to 205%).
DOI 原文 ·
@inproceedings{paperbot8443,
title = {ART: A Unified Unsupervised Framework for Incident Management in Microservice Systems},
author = {Yongqian Sun and Binpeng Shi and Mingyu Mao and Minghua Ma and Sibo Xia and Shenglin Zhang and Dan Pei},
booktitle = {ASE},
year = {2024},
doi = {10.1145/3691620.3695495}
}