Big names launch open source SDN project
Apr 8, 2013
A slew of incumbent network equipment-makers has got together to agree a Software Defined Networks (SDN) approach called OpenDaylight, it's been announced.
All the big US network vendors are on-board and the group says the resulting open technology will be designed to play nice with OpenFlow. At this point OpenDaylight is supported by the likes of Cisco, Juniper, HP, Nuage Networks (Alcatel-Lucent), IBM, VMware and Microsoft. Ericsson, NEC and Fujitsu are there too, Huawei, unsurprisingly, is not.
The SDN start-ups and the big SDN/OpenFlow proponents like Google and Facebook are not part of the effort at this stage either.
On the face of it, then, OpenDaylight appears to be a classic disrupt the disruption move.
Expanation: It used to be that incumbent industries behaved in a set pattern when in the process of being disrupted, all according to the Clayton Christensen (Innovator's Dilemma) script: first they didn't see it, when they did they denied, then they tried to undermine, when that failed they got beaten and disappeared.
Today in IT, the incumbent companies tend (indeed have been management trained) to see the disruptions coming and get clever.
Nobody doubts that SDN has an important role for all the data-heavy users - media comanies, ISPs, OSPs, telcos, data centres and cloud service providers. SDN is the software smarts that will control the 'separated' data and management planes in the next generation network.
With current generation networks, data forwarding and the software that controls it are generally mixed up together in specific, high-powered (and expensive) boxes - routers, switches, gateways and the like.
The promise of SDN is that powerful 'standard' hardware can provide the network horsepower with centralised software control, thus enabling telcos/ISPs to cope with the data deluge expected in the next ten years or so without breaking the bank.
The problem is that all the incumbent equipment vendors are heavily invested (both financially and intellectually) in the old box approach. As SDN is just inevitable, the clever thing to do is attenuate the impact of SDN by making sure it's compatible and adds some value to the existing and future boxes of the traditional equipment vendors.
That way SDN becomes an overlay on the existing network, squeezing some extra utility out of the incumbent equipment - rather than a harbinger of doom for the box approach.
Whether this is the aim of OpenDaylight or not should become apparent in the next few months.