paperbot · PL 论文追踪

全部论文 待读 收藏

P2 TraceDiag: Adaptive, Interpretable, and Efficient Root Cause Analysis on Large-Scale Microservice Systems

TraceDiag：面向大规模微服务系统的自适应、可解释且高效的根因分析
FSE Proceedings of the 31st ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering2023匹配度 2/10 · 该论文关注大规模微服务系统的运行时根因分析，使用强化学习剪枝服务依赖图，属于分布式系统与AIOps/调试领域；它与用户核心方向（程序静态分析、LLM/Agent 与程序分析交叉、浮点程序分析、贝叶斯程序分析）基本无关，仅在“调试与修复/工业落地”上有很弱的关联。引用 44 inbox
Ruomeng Ding, Chaoyun Zhang, Lu Wang, Yong Xu, Minghua Ma, Xiaomin Wu, Meng Zhang, Qingjun Chen, Xin Gao, Xuedong Gao, Hao Fan, Saravan Rajmohan, Qingwei Lin, Dongmei Zhang
测试、调试与修复分布式与容错
crossref: 2026-09-12dblp: 2026-09-12arxiv: 2026-09-12

AI 速览

一句话：提出端到端 RCA 框架 TraceDiag，用强化学习剪枝服务依赖图，提升大规模微服务根因分析的效率与可解释性。

现代微服务系统通常包含数百个组件，根因分析（RCA）高度依赖人工且效率低下。TraceDiag 是面向大规模微服务系统的端到端 RCA 框架：它利用强化学习为服务依赖图学习剪枝策略，自动去除冗余组件，且策略具有可解释性并能自适应新的 RCA 实例；随后在剪枝后的图上执行基于因果的 RCA 方法，从而兼顾准确率与效率。论文在 Microsoft Exchange 的真实 trace 上评估，性能优于当前先进 RCA 方法；TraceDiag 还集成到 Microsoft M365 Exchange，显著提升系统可靠性并减少 RCA 人工投入。

问题
大规模微服务系统组件众多，根因分析（RCA）耗时耗力，现有方法难以同时保证效率与准确性。
方法
TraceDiag 用强化学习为服务依赖图学习可解释、自适应的剪枝策略，去除冗余组件，再在剪枝图上执行因果 RCA。
评估
在 Microsoft Exchange 真实 trace 上对比先进 RCA 方法，并已在 Microsoft M365 Exchange 集成部署，验证了系统可靠性与人工成本方面的收益。
原文摘要（Abstract）

Root Cause Analysis (RCA) is becoming increasingly crucial for ensuring the reliability of microservice systems. However, performing RCA on modern microservice systems can be challenging due to their large scale, as they usually comprise hundreds of components, leading significant human effort. This paper proposes TraceDiag, an end-to-end RCA framework that addresses the challenges for large-scale microservice systems. It leverages reinforcement learning to learn a pruning policy for the service dependency graph to automatically eliminates redundant components, thereby significantly improving the RCA efficiency. The learned pruning policy is interpretable and fully adaptive to new RCA instances. With the pruned graph, a causal-based method can be executed with high accuracy and efficiency. The proposed TraceDiag framework is evaluated on real data traces collected from the Microsoft Exchange system, and demonstrates superior performance compared to state-of-the-art RCA approaches. Notably, TraceDiag has been integrated as a critical component in the Microsoft M365 Exchange, resulting in a significant improvement in the system's reliability and a considerable reduction in the human effort required for RCA.

链接与引用

DOI 原文 · arXiv · PDF（开放获取） · DBLP

BibTeX 
@inproceedings{abs-2310-18740,
  title = {TraceDiag: Adaptive, Interpretable, and Efficient Root Cause Analysis on Large-Scale Microservice Systems},
  author = {Ruomeng Ding and Chaoyun Zhang and Lu Wang and Yong Xu and Minghua Ma and Xiaomin Wu and Meng Zhang and Qingjun Chen and Xin Gao and Xuedong Gao and Hao Fan and Saravan Rajmohan and Qingwei Lin and Dongmei Zhang},
  booktitle = {FSE},
  year = {2023},
  doi = {10.1145/3611643.3613864}
}