Efficient Bare Metal Auto-scaling For NFV Platform

There are always a cloud data center behind the Network Function Virtualisation(NFV)system.Typically,elasticity is an essential attribute of cloud data center,which is critical for operating resources in face of peaks and valleys of business.At present,the automatic scaling technique of virtual machines is widely studied,but barely for physical machines.Despite lack of flexibility,we all know that physical server can perform faster and more efficiently than virtualized instances,especially in NFV systems.Some virtual network functions(VNFs)actually require high performance computing,which is a hard task for virtual machines.Deep packet inspection,for example,is now based on customized hardware and running these functions as a hypervisor on a commercial server can result in 10x performance degradation.Based on the above scenario,this paper proposes a scheme based on NFV MANO architecture and supporting the automatic scaling of bare metal resources.A great deal of integration has been made to the automatic management of bare metal.The automatic discovery and reporting of new bare metal is introduced.Automated deployment of virtual network functions and the introduction of a bare metal resource pool similar to that of a virtual machine,the release of bare metal resources and Assign to do better management.This topic selects the Zabbix monitoring platform,realizes the real-time monitoring to bare metal and the function of the virtual network carried by it.And developed an automatic scaling algorithm for bare metal resources.The algorithm based on load forecasting,combined with the bare metal processing flow to set the cache state,can efficiently handle the bare metal resource-related scheduling tasks.And finally make tests on an open source NFV platform.The new scheme obtains good result on computation intensive VNFs scenario,including complete the scale in half an hour,guarantee for the continuity of VNF processing business,and can cope with the load fluctuation better.Besides,good management of bare metal resources can be significant for the data center power cost and human maintenance cost.Accordingly,we think that auto-scaling of physical machine is worth studying.