Standards war on the cards?

The news that the Open Networking Foundation (ONF) will focus on software-defined networking (SDN) as 'the' technology to take wireless and mobile networking to the next level, broke early in 2014.

For SDN advocates, this was immediately welcomed. As Serge Manning, head of the ONF's wireless and mobile working group is reported to have said, there is a need for protocol extensions, new parameters "and slightly different ways of doing things that will bring more value to wireless networks".

Of course, the ONF is an organisation dedicated to the promotion and adoption of SDN through open standards development, so its stance on SDN is to be expected.

Does this invalidate SDN or devalue its position? On the contrary, SDN is increasingly seen by many vendors as a cost-effective, adaptable answer to demands by wired, wireless and mobile network users for more programmability, tighter control and broader-based automation.

One of the more significant benefits of SDN - which decouples networking control from the data-forwarding function - is its ability to allow administrators to integrate best-of-breed technology into their networks. In this light, SDN will become much more 'deployable', evolving to address the practical demands of businesses for more dynamic, agile networks capable of deploying new services through common APIs.

By-products will be greater simplicity in wireless network architectures, better security, and - most importantly - increased reliability, which will bring with it undeniable operating expenditure savings.

There is a close relationship between the OpenFlow standard and SDN. It's to be expected, as the ONF's "signature accomplishment" (its words) has been the introduction of OpenFlow.

What is OpenFlow? It's an open standards-based communications protocol added as a feature to many commercial Ethernet switches, routers and wireless access points. One of its achievements has been to move network control out of proprietary network switches and into control software that's open source and locally managed.

OpenFlow also provides a standardised 'hook', allowing vendors to run experiments without requiring them to expose the internal workings of their network devices. As it is the first SDN standard, it's seen by many as a key element in future SDN-oriented network architectures.

Will its position be challenged by the latest open standards development from the OpenDaylight Project? This is a collaborative, vendor-led initiative under the banner of the Linux Foundation, which is committed to foster innovation and create an open and transparent approach to SDN.

Hydrogen is the first release by the OpenDaylight Project team. It is a modular, open source SDN platform featuring an architecture designed to empower multiple-use solutions in network and cloud service providers - as well as enterprise IT environments. Hydrogen was designed using open source development best practices with input from a number of leading IT organisations.

It comes in three flavours, the first of which is the Hydrogen Base Edition, an SDN starter kit which is said to be useful for proofs of concept, particularly for OpenFlow networking. Next up is Hydrogen Virtualisation Edition, which includes several modular enhancements enabling the creation and operation of virtual network overlays in data centres. The third option is Hydrogen Service Provider Edition, essentially an expanded version of Base Edition.

There is a close relationship between the OpenFlow standard and SDN.

In addition to facilitating third-party contributions - more than a million lines of code authored by over 150 developers landed in the Hydrogen release - it ships with software that enables traffic engineering employing routing protocol library input, together with topology and path programming models. These are aimed at exchanging routing information between gateway hosts in a network of autonomous systems such as those which characterise the Internet.

Based on these developments, will a standards war break out between the ONF and OpenDaylight? Will there be a fight for the 'soul' of SDN?

This is unlikely, as OpenDaylight includes support for the OpenFlow protocol in line with a broad, industry-wide understanding across vendors, service providers, end-customers and researchers who agree that, while OpenFlow is a useful protocol in many scenarios, SDN is not limited to OpenFlow or any single protocol, for that matter.

With software - via SDN - playing a more significant role in networking, the emphasis should be on the concept of open source (rather than standardisation) as it was back in the days of Linus Torvalds and the development of the Linux kernel, one of the most prominent examples of free and open source software collaboration.

Today, more than 95% of the world's 500 fastest supercomputers run some variant of Linux, including those powering the Internet, Amazon, Facebook and many other platforms.

The success of Linux is a clear example of the power of open source versus standardisation. That said, perhaps there is a role for both standards and open source within SDN. Ideally they should be within the context of a streamlined and automated software ecosystem cable of managing ever more devices. This will go a long way towards meeting the needs of IT users, currently crying out for cuts in corporate-wide network operating expenditure.