摘要
The varied network performance in the cloud hurts application performance.This increases the tenant’s cost and becomes the key hindrance to cloud adoption.It is because virtual machines(VMs)belonging to one tenant can reside in multiple physical servers and communication interference across tenants occasionally occurs when encountering network congestion.In order to prevent such unpredictability,it is critical for cloud providers to offer the guaranteed network performance at tenant level.Such a critical issue has drawn increasing attention in both academia and industry.Many elaborate mechanisms are proposed to provide guaranteed network performance,such as guaranteed bandwidth or bounded message delay across tenants.However,due to the intrinsic complexities and limited capabilities of commodity hardware,the deployment of these mechanisms still faces great challenges in current cloud datacenters.Moreover,with the rapid development of new technologies,there are new opportunities to improve the performance of existing works,but these possibilities are not under full discussion yet.Therefore,in this paper,we survey the latest development of the network performance guarantee approaches and summarize them based on their features.Then,we explore and discuss the possibilities of using emerging technologies as knobs to upgrade the performance or overcome the inherent shortcomings of existing advances.We hope this article will help readers quickly Received:Apr.07,2020 Revised:Oct.23,2020 Editor:Haifeng Zheng understand the causes of the problems and serve as a guide to motivate researchers to develop innovative algorithms and frameworks.
基金
This project is partially supported by the National Natural Science Foundation of China(No.61872401)
Fok Ying Tung Education Foundation(No.171059).