Mike Fratto, my former editor at Network Computing, now a consultant at Current Analysis, recently wrote a blog entry that tried to cut through the 3 Hs of SDN — Hype, Hoopla and Hysteria — by posing what some consider the essential question of life (unless you’re a Monty Python or galactic hitchhiker fan, in which case the answers are ‘be nice’ and ’42’, respectively), “What’s in it for me?”
“In the drive to define SDN, established and start-up networking vendors are developing products that can improve your network operations, and that is what is important. […] What matters are not the foundational ideas underpinning a new technology, but the practical applications,” Fratto said.
While a sizable number of the constituents of the IT industry — builders, sellers, buyers, and opinion-makers — run around like headless chickens squawking about what is and isn’t SDN, real people are solving real problems with real solutions. It started several years ago with academia and the biggest network users like eBay, Google, Amazon and Facebook scrambling for solutions as they came up against the massive scaling involved with the Wild West Web. Next came the carriers and service providers. And eventually, enterprises, public sector, and smaller businesses will get involved, once most of the bugs have been worked out, and costs and complexities have been made more palatable.
Various analysts have sized current and projected SDN markets, and with surging network demand coupled with growing network obsolescence, the sums involved are significant. IDC recently reported that the enterprise and cloud service provider segments are forecast to explode more than tenfold, from $360 million in 2013 to $3.7 billion in 2016.
For the incumbent network vendors — mainly Cisco — SDN is a huge threat. FBR Capital Markets sees SDN adoption driving sales of a commodity, low-margin switches at the expense of Cisco/Juniper’s conventional switch/routers. FBR analyst Scott Thompson forecasts that there will be a 40% drop in switch/router ports deployed by service providers/large enterprises in the next 18-36 months.
FBR believes that the next generation of network and data center hardware will blur the boundaries between routing, switching, and computing, providing a single hardware platform on which network and computing functions will be delivered through software applications.
Here are FBR’s 9 characteristics driving the evolution of technology:
- drive power efficiencies from a total platform perspective;
- lower the overall data center footprint;
- seek to eliminate redundant or nonessential hardware and components;
- commoditize hardware;
- drive the commoditization of hardware through the scaled use of non-branded component-based hardware solutions in distributed datacenter architectures;
- attempt to replace ASICs with “open” merchant silicon and/or drive advanced functionality into general purpose CPUs;
- seek to utilize and optimize “open source” software and hardware alternatives; and,
- increase infrastructure flexibility through a software approach.
The transition to SDN and the next phase of networking won’t be fast or without its share of problems.
Some of the vendors are apparently trying to ‘control’ the evolution of SDN via their own standards groups, according to GigaOm. Last month, the nonprofit Object Management Group (OMG) said Dell intends to create an OMG committee on SDN. This proposal followed moves by Cisco, Citrix, HP, and other vendors to sponsor or contribute to the consortium which led to the freshly-announced OpenDaylight initiative.
What isn’t in doubt is that network professionals and organizations will continue to seek solutions that address their needs, and vendors can either help them, or get left behind.