论文信息 - Achieve high availability about point-single failures in OpenStack

Achieve high availability about point-single failures in OpenStack

This paper describes a method to solve the complexity of distributed environment in cloud technology and the single-point failure problem. In this paper, we take the virtual machine failure and host failure in OpenStack into consideration. The ability of fast restoration of this service is achieved by components in OpenStack which is called ceilometer and new components named Senlin. The function is achieved by adding patches and modifying the source code in OpenStack. Ceilometer plays a role of collecting information of the virtual machine failure and host failure in OpenStack and transfers this information to Senlin. Then Senlin can make a decision to restore the failure. This clever method achieves the high availability of OpenStack. We implement the proposed method and show the effectiveness of fast restoration.

Xiaoyong Li | Yiping Wang

[1] Francis Tam. Service Availability Standards for Carrier-Grade Platforms: Creation and Deployment in Mobile Networks , 2009 .

[2] Yiming Li,et al. Software defined networking: State of the art and research challenges , 2014, Comput. Networks.

[3] David E. Irwin,et al. Virtual Machine Hosting for Networked Clusters: Building the Foundations for "Autonomic" Orchestration , 2006, First International Workshop on Virtualization Technology in Distributed Computing (VTDC 2006).

[4] Abdelkader H. Ouda,et al. Resource allocation in a network-based cloud computing environment: design challenges , 2013, IEEE Communications Magazine.

[5] Luiz André Barroso,et al. The tail at scale , 2013, CACM.

[6] Adam Wierman,et al. The Economics of the Cloud , 2017, ACM Trans. Model. Perform. Evaluation Comput. Syst..

[7] Abdallah Shami,et al. NFV: state of the art, challenges, and implementation in next generation mobile networks (vEPC) , 2014, IEEE Network.

[8] Calton Pu,et al. Improving Performance and Availability of Services Hosted on IaaS Clouds with Structural Constraint-Aware Virtual Machine Placement , 2011, 2011 IEEE International Conference on Services Computing.