V2PFQL: A proactive fault tolerance approach for cloud-hosted applications in cloud computing environment

Abstract

Proactive Fault tolerant (FT) approaches for infrastructure's cloud computing which have been increasingly developed in research scope with the support of machine learning need consistent monitoring and highly accurate predicting systems. However, it is difficult to implement the proactive fault tolerant approach of infrastructure's cloud computing because of complex architecture, dynamic services, and many interdependent factors existing in cloud environment. To overcome this problem, a proactive FT approach for cloud-hosted application is proposed to achieve a high degree of automation, including physical machine (PM) fault analysis and self-learning virtual machine (VM) migration. It is based on MAPE-K (Monitor-Analyze-Plan-Execute over a shared Knowledge) loop to take into account the knowledge inside controllers in run time. To evaluate PM fault analysis model, the web e-Commerce is deployed by OpenStack Stein while the prototype of VM migration component is built by CloudSim for VM migration solution evaluation. The effectiveness of the approach is demonstrated in experiment results.