I loved reading “Accelerating NFV Delivery with OpenStack,” a white paper recently published by the OpenStack Foundation that documents the strides OpenStack is making. In collaboration with other open source projects such as OPNFV, standards bodies such as ETSI, and international telecom consortia, it is moving to become an ideal cloud platform for hosting NFV workloads.
The idea of using OpenStack as a platform for network functions virtualization (NFV) is pretty well established and has been a hot topic at OpenStack Summits for the last year and a half. The essentials are simple.
Communication service providers need OpenStack to help them move into the cloud era so they can finally retire proprietary, hardware-bound service architectures in favor of logically chained, scalable, software-managed virtual network functions (VNFs) running on OpenStack. This migration will also embrace increasingly commoditized compute, storage, and network infrastructures. And, because telecoms own the pipes, they have an added business imperative to build their own public cloud options.
Like enterprise and Web/SaaS adopters, they want to use OpenStack to commoditize their infrastructures, pushing down costs for what they need the most. They want to pay less for software and hardware. They want to automate operations and enable customer self-service. They want to speed up and standardize new-service development so they can compete more effectively with greenfield upstarts (for example, WhatsApp) and stake out new and profitable shares of mobile, IoT, and other fast-growing communications markets.
It’s All About Agility
Today, a multitude of open source projects and communities are coming together to enable NFV’s true value proposition: service agility. Service agility is a pressing business need for communication providers today to meet the exploding global demand for digital and mobile services. It’s exciting that OpenStack is at the nucleus of it all.
OPNFV is helping bring these various open source projects and communities together using continuous integration (CI) frameworks. OPNFV’s upcoming Release B (Brahmaputra) brings this effort to fruition for the second time. Arno, the first release last year, set the stage to bring together OpenStack, KVM, DPDK, Open DayLight, and more. With the OpenStack Fuel project as a lead installer for OPNFV, symbiotic collaboration between OpenStack and OPNFV is well on its way.
Because of this, NFV has become a poster child for what OpenStack can do.
OpenStack is where the stacks, VNFs, and management software run; where the SDN connects; and where the specialized components integrate. And, as with every other use-case, providing an open framework by design is OpenStack’s greatest strength.
While we normally avoid the too-often carelessly applied analogy that “OpenStack is an operating system for the cloud,” the analogy holds true in at least one key respect: that the great strength of standard frameworks is their supreme adaptability.
Nobody blinks to think that Linux, for example, can run a small network router or a high-end 3D graphics workstation (with appropriate drivers and graphics cards). Nor should anyone wonder whether OpenStack can, in principle, frame clouds that run cloud-native microservice applications, or distributed NFV platforms for telecommunications carriers.
A Model for Enterprises
What’s more, NFV isn’t even exclusive to telecoms. Although the initial use-cases (such as vCPE) typically cite telecoms, the advantages of NFV and the fine-grained control it brings in the way of VNFs are just as relevant to enterprises. VNFs such as virtual load balancers, virtual routers, and virtual firewalls are equally useful and relevant to enterprises whether they’re financial service providers that need to reduce network latency or complex enterprises that need granular control of access to virtual machines.
With all the attention NFV has been getting, it’s important to remember that the telecom industry only represents about 12% of OpenStack’s user base. Enterprises across financial services, media, automotive, Web/SaaS, and many other sectors are using OpenStack because of the agility it offers their businesses.
I urge enterprises to watch what telecom carriers are able to do using OpenStack and think about what OpenStack’s flexibility and agility can do for their businesses. The efforts going into OpenStack to increase flexibility, scale, and performance benefit all.