This is a paper published on CCS’ 2017 by members from Feifei Li’s group. The paper mainly focus on the log analysis of large scale system, for the purpose of debug.
In real world, we debug a system by analyzing the log records produced by different sources from the system. For example, the network log and the storage log in an openstack service. In this paper, the authors combined the deep learning model with log analysis, to be more specific, the model utilize the RNN to catch the characters between timestamps.