Heuristic-based Parsing System for Big Data Log

Logs are essential for recording system runtime information and are widely used for effective service management. However, traditional log parsing methods struggle with scale, leading to inefficiencies and reduced model accuracy. 

‍

This paper presents Aclog, an automated tool designed for large-scale log analysis, storage, and auditing. Aclog structures unstructured logs into a unified format. Its effectiveness is demonstrated using a large-scale public log dataset from ten widely used systems.

‍