The successful adoption of NFV depends on the ability of a virtualized cloud infrastructure to deliver sufficient performance comparable to today’s proprietary network appliances.
Enterprise/IT cloud hardware specifications and configurations are not optimized to handle heavy NFV dataplane workloads needed by Communications Service Providers. In order to achieve this kind of performance, you need advanced processor and networking technologies embedded in silicon and in PCIe devices, but even so, you need to ensure your VNFs are ultimately able to leverage these capabilities.
In this talk, we'll show how NFV throughput performance varies depending on proper VNF workload placement configuration, by declaring what the VNF requires from the hardware (SR-IOV, DPDK, etc) and ensuring proper matching and configuration in NFVI. We will see how Enhanced Platform Awareness and intelligent orchestration are enablers for such extreme performance.