Astragalus: Automatic Configuration Repair for Production Networks

2026-06-12 · Source: cs.SE updates on arXiv.org · Field: Technology & Digital — Software Development & Engineering, Cloud Computing & IT Infrastructure, Artificial Intelligence & Machine Learning · Depth: Expert, extended

Summary

Astragalus is an automatic configuration repair (ACR) tool designed to address network misconfigurations, a major cause of service outages. Unlike existing "semantic-driven" approaches that struggle with scalability due to complex SMT constraints, Astragalus employs a "syntax-driven" method inspired by automatic program repair. It utilizes a "localize-fix-validate" pipeline to efficiently identify and correct errors. The tool demonstrated high effectiveness, repairing 100% of incidents in synthesized networks and 97.5% in a real network with 15 types of injected errors, averaging 7.36 seconds per repair. It also provided valid suggestions within 6 minutes for four recent incidents in a production network of O(1,000)-O(10,000) devices, proving significantly faster and more scalable than prior solutions like AED and CEL.

Key takeaway

For network operators managing large-scale production networks and struggling with misconfiguration-induced outages, you should evaluate "syntax-driven" automatic configuration repair (ACR) tools like Astragalus. This approach significantly accelerates fault localization and repair, often resolving incidents in seconds, far surpassing the scalability of "semantic-driven" methods. While not every complex root cause is directly identified, these tools provide actionable suggestions that drastically reduce manual troubleshooting time, improving network stability and operational efficiency.

Key insights

Syntax-driven automatic configuration repair offers superior scalability and generality over semantic-driven methods.

Principles

• Network configurations exhibit "plastic surgery hypothesis" due to role redundancy.
• Fast network verification enables efficient validation of numerous candidate fixes.
• Syntax-driven repair generalizes across diverse protocols and vendors.

Method

Astragalus employs a "localize-fix-validate" pipeline: localize suspicious lines via SBFL, generate candidate fixes (remove, insert, modify) from existing configurations, then validate using network verifiers.

In practice

• Apply Ochiai as the default SBFL technique for localization.
• Prioritize "Remove" operations when generating configuration fixes.
• Integrate existing network verifiers for rapid candidate fix validation.

Topics

• Automatic Configuration Repair
• Network Configuration
• Syntax-driven Repair
• Fault Localization
• Network Verification
• Data Center Networks

Best for: Research Scientist, IT Professional, Operations Professional

Related on AIssential

• Evaluating Agentic Configuration Repair for Computer Networks — Agentic LLM architectures, augmented with verification and context tools, significantly improve network misconfiguration repair.
• PropLLM: Propagation-Aware Scene Reconstruction for Network Fault Diagnosis — PropLLM uses LLMs with hop-by-hop scene reconstruction and causal attention for accurate network fault diagnosis, reducing ambiguity and hallucination.
• asyncinject 0.7 — AI models can proactively identify and fix bugs in software dependencies, enhancing library maintenance.
• STMutants: A Mutation Testing Dataset for Structured Text Programs in Industrial Automation — STMutants provides the first public mutation testing benchmark for IEC 61131-3 Structured Text, enabling reproducible research and LLM evaluation for PLC software.
• STAR: A Stage-attributed Triage and Repair framework for RCA Agents in Microservices — STAR improves LLM-based RCA agent reliability by localizing and repairing errors within specific diagnostic stages.
• Article: When a Cloud Region Fails: Rethinking High Availability in a Geopolitically Unstable World — Traditional cloud high availability models, relying on multi-AZ deployments, are insufficient for geopolitical disruptions, which can compromise entire regions as correlated units due to events like s…

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.SE updates on arXiv.org.