An Approach to Pinpointing Bug-Induced Failure in Logs of Open Cloud Platforms

Authors:
  • JTJia Tong
    Jia Tong
  • LYLi Ying
    Li Ying
  • THTang Hongyan
    Tang Hongyan
  • WZWu Zhonghai
    Wu Zhonghai

Abstract
Software bugs have been one of the dominant causes of system failures, especially in cloud systems based on open source platforms. One big challenge for troubleshooting these cloud systems is to pinpoint the software bug-induced failure in large and complex log files which is a nightmare for administrators. So far, there has been little study on how to identity bug-induced failures based on log analysis. In this paper, we analyze and describe features of bug-induced failure logs from bug repository and Q&A websites, and then propose a general automatic approach to pinpoint logs of bug-induced failure from log files of open cloud platform. In the approach, two algorithms called MPIN and SPIN are presented for log classification. We evaluate our approach by applying logs collected from bug repositories of OpenStack and Hadoop, and five Q&A websites. The experimental result shows that the proposed approach can identify logs of bug-induced failure in OpenStack logs with 83.9% precision, and for Hadoop logs with 82.52% precision.

This publication has 10 references indexed in Scilit: