Structure Design of Log Management Platform That Supports Smooth Expansion of Anomaly Detection Capability
Honglu Gan, Haoran Du, Yulu Cao · 2022
With the popularity of cloud services and more and more traditional businesses on the cloud, the log management platform as an important cloud service is suffering from the impact of more new anomaly detection tasks. Traditional log management platforms have some problems in anomaly detection, such as high maintenance costs and inflexible detection tools. This paper proposes an architecture of a log management platform that supports the smooth expansion of anomaly detection capability. This architecture allows users to upload the anomaly detection model, and the online operation of the model has been realized through the model online loading module, to achieve the smooth expansion of the anomaly detection capability.