High Load Function Prediction Model Based on Decision Tree
Qian Lin, Hengmao Pang, Jun Yu, Guangxin Zhu, Mingjie Xu, Lin Wang, Haiyang Chen · 2019
Cloud computing platform application for adaptive non-downtime upgrade has become one of the current hot research issues. However, the system self-adapting upgrade of high-load function will bring some risks, so it is necessary to upgrade the high-load function under the supervision of the relevant technicians as far as possible. Many of the load values of cloud computing platform applications are dynamic and related to many factors. Current functional awareness can only determine the load of the current function, and can not predict the load when upgrading. Therefore, an algorithm is urgently needed to determine the load level of upgrade according to a variety of factors. In this paper, a decision tree based prediction model is proposed, which can predict whether the load is high when a particular function is upgraded by the state value of the current system.