OpsMem: Dual-Memory Reasoning with Cross-Memory Resonance for Failure Diagnosis

2026-07-13

AI Summary

This work addresses the lack of synergy between diagnostic states and operational experience in modern software fault diagnosis by proposing OpsMem, a dual-memory framework. OpsMem maintains the current diagnostic state in short-term memory while storing reusable operational knowledge in long-term memory. A cross-memory resonance mechanism dynamically retrieves and activates relevant historical experiences to guide large language model–driven multi-agent collaborative diagnosis, with newly acquired insights continuously consolidated back into long-term memory. This approach uniquely integrates dual-memory architecture with cross-memory retrieval, enabling dynamic co-evolution between the diagnostic process and the experience repository. Evaluated on a real-world microservice failure dataset from Huawei, OpsMem achieves significant improvements of 46.88% in Match and 18.39% in Relevant metrics over existing baselines.

Abstract

Failure diagnosis in modern software systems requires iterative evidence acquisition and hypothesis reasoning guided by operational experience. Existing LLM-based methods improve diagnosis through agentic reasoning or knowledge augmentation, but they often lack a mechanism to coordinate the evolving diagnostic state with operational experience during iterative diagnosis. We propose OpsMem, a dual-memory framework that maintains a short-term memory for the current diagnostic state and a long-term memory for reusable operational experience. OpsMem uses cross-memory resonance to activate state-relevant long-term memory, conditions multi-agent diagnosis on the short-term and activated long-term memories, and consolidates reusable experience from solved incidents back into long-term memory. Experiments on a real-world Huawei microservice failure diagnosis dataset show that OpsMem outperforms representative agentic-reasoning and knowledge-augmented baselines, improving Match and Relevant by up to 46.88% and 18.39% over the strongest baseline, respectively.

Problem

• failure diagnosis
• operational experience
• iterative reasoning
• diagnostic state
• memory coordination

Innovation

• dual-memory reasoning
• cross-memory resonance
• failure diagnosis
• operational experience
• multi-agent reasoning