An learning-based fault-tolerant model for real-time applications on clouds
Haoran Han, Weidong Bao, Xiaomin Zhu, Xiaosheng Feng · 2017
The rapid development of mobile technology has further promoted the scale of real-time applications, such as mobile payment, real-time positioning and mobile communication. At the same time, fault-tolerant requirements of data and process for applications increasingly emerged. Complexity and particularity of the traditional fault-tolerant mechanisms in cloud can't meet the modern fault-tolerant requirements and consume a lot of cloud resources. To solve this problem, we propose an efficient learning-based fault-tolerant model, Nebula, Nebula not only ensuring the fault tolerance of real-time application, but also greatly improving the utilization of cloud resources.