6:00 AM -- Software-defined networking (SDN) rapidly expanded to take up nearly all of tech-industry journalism this year, at least in the networking sphere. Rather than make a year-end list out of it, I thought I'd share a few of the trends I'll be watching for.
For a more granular look at the subject, check out the SDN questions recently posed by Heavy Reading analyst Graham Finnie. I'll be watching the technology at the nuts-and-bolts level as well, but I kept this list rather high-level, looking mostly at business implications.
Where did I miss the mark? Talk it out on the message boards below. SDN is still morphing, so there's plenty to debate.
More startups get acquired
Might as well get this out of the way first, because it's so obvious. VMware Inc.'s $1.2 billion acquisition of Nicira was atypical in terms of the money involved, but there's no denying that an SDN feeding frenzy is underway. Brocade Communications Systems Inc., Cisco Systems Inc. and Juniper Networks Inc.
have since made acquisitions. More are likely to come, maybe from the names listed below.
Big players move closer together
VMware is becoming more like Cisco, and Cisco more like VMware parent EMC Corp. (note Cisco's apparent funding of storage startup WhipTail, as reported by TechCrunch). Oracle Corp. owns another hardware company in Xsigo Systems Inc., its not-really-SDN acquisition.
These giants are becoming more similar with every acquisition. It's no stretch to say VMware has become a networking company, and of course Cisco has trodden server vendors' turf with its Unified Computing System (UCS). In the coming year, the points of conflict among these companies will only grow.
It's not directly an "SDN" issue; in fact, you could say it has more to do with the cloud. In any event, it's not going to stop.
Carriers use cases emerge
Carriers really are using SDN. We've heard Verizon talk about OpenFlow, and some of Cyan's rural customers are experimenting with SDN-like monitoring on the fiber network. Lots of ad-hoc uses are emerging, too; we'll be running an interview after the holidays where David Meyer of Brocade briefly discusses that.
They're just dabbling, in many cases, but it's work that could help solidify the shape of carrier SDN. It will be worth watching.
Standards will overrun us all
Certainly, SDN will need more than just OpenFlow and the Open Networking Foundation. But now there's also network functions virtualization (NFV) being pitched by carriers, and various SDN work brewing at the Internet Engineering Task Force (IETF), Institute of Electrical and Electronics Engineers Inc. (IEEE), Alliance for Telecommunications Industry Solutions (ATIS) and International Telecommunication Union (ITU).
I don't have much of a point here, other than to complain about the difficulty of following all this.
Commoditization won't go away
The argument is that SDN can replace switches, and possibly routers, with generic and inexpensive boxes controlled externally. The counter-argument is that SDN will instead add a richness and a programmable intelligence to switches and routers, something the vendors will be able to exploit.
SDN won't drive Cisco and its rivals out of business in 2013. But the carriers have heard the commodity argument, and they're intrigued. Any sign that it's becoming a practical reality would be bad news for vendors, especially Cisco.
Don't expect the revolution right away. But watch for those signs.
— Craig Matsumoto, Managing Editor, Light Reading