Oct 13, 2014
The MEF (Metro Ethernet Forum) is arguably the best-functioning global industry networking forum. So when Bob Metcalf (the inventor of the Ethernet standard) on behalf of MEF launches a vision of Networking-as-a-Service — dubbed the ‘Third Network’ — to be realized on top of Carrier Ethernet 2.0, we sit up and take note.
The central element of the Third Network is to provide QoS over the Internet, combining Private and VPN based on CE (Carrier Ethernet) 2.0 with the Internet. CE 2.0 connections use transport technologies such as Ethernet, MPLS, and OTN (Optical Transport Network), delivering services with assured performance and security; however, these connections require time and effort to be initiated across network operator domains. The Internet delivers on-demand, ubiquitous services, but promises only best-effort connections, leaving users with changing security and performance levels.
MEF’s Third Network envisages a worldwide Network-as-a-Service built on SDN/NFV (Software-Defined Networks and Network Functions Virtualization). The aim is to create agile (dynamic, on-demand services), assured (delivering performance and security assurance) and orchestrated (automating service lifecycle management within and across network operator domain) services. NFV provides the general purpose VMs for network functions, while SDN defines network control functions in central controllers. The first step in the MEF vision is to develop standardized APIs to provide the abstraction of the different technology layers. Proof of Concept demonstrations are scheduled for the Global Ethernet Networking 2014 event in November in Washington.
The ramifications of enterprise-level QoS services on the public Internet are significant and are central to the whole Net Neutrality debate. The increased mobility of the workforce and the fast uptake of working-from-home increases the use of latency-sensitive services like video conferencing and ERP that require better QoS connectivity options. This may be something that business users will be willing to pay for, which can create new revenue streams for carriers, ISPs, and software developers.
However, there is no such thing as a Free Lunch, and QoS for one user on a shared network means less bandwidth for other users — thus challenging the central tenet of Net Neutrality: that all users have the same right to Internet access, and no one should be able to buy higher service levels at other users’ expense. But how far do Internet access rights extend? How much bandwidth is ‘enough bandwidth’? Should we distinguish between major OTT (over-the-top) video bandwidth-huggers like Netflix, and enterprise users looking for QoS from home? The resolution of these issues in Europe may well define the level of infrastructure investment over the next decade, as telecom operators say that charging for different speeds and services would help them invest in upgrading their networks, and allow European infrastructure investments to catch up with the US and Asia.
One indication of the near-future deployment of MEF’s Third Network came from the recent EU Parliament grilling of the new EU Commissioner for the Digital Economy and Society, Germany’s Günther Öttinger. He said that Net neutrality “is a common interest for all users and for all citizens […] there may be an exemption, but not for businesses, not for business cases, therefore we need neutrality for all users”. Such ’exemptions’ will likely not allow Netflix and other content providers to buy better QoS for their program delivery, but it leaves room for QoS services to individual users connecting to the public Internet. Allowing them to buy QoS bandwidth — typically when using video conferencing apps — supports improved European competitiveness.
So we are back to apps as the real revenue generator! The Third Network Vision will allow infrastructure providers, ISPs and apps developers to monetize capex investments faster, and perhaps attract more external investors.
Image credit: Jason Walton (flickr) / CC-BY