& cplSiteName &

SDN: Start Making Sense

Heavy Lifting Analyst Notes
Heavy Lifting Analyst Notes
Heavy Lifting Analyst Notes
8/13/2012
50%
50%

Very few – if any – days have gone by over the past few months in which the relative merits of software-defined networking (SDN) have not been publicly debated.

As we note in the latest Heavy Reading report, "SDN & the Future of the Telecom Ecosystem," the debate is likely to increase in the near future given the potential impact for both carriers and vendors.

Conceptually, SDN and the approach of separating the control plane from the data plane present a strong value proposition. And while this approach has long been adopted in other parts of the network (session border controllers and IMS core), it does hold the promise to revolutionize and break new ground on the way data transport networks function.

Here's a quick take on the factors and unanswered questions that ultimately will determine if SDN becomes a truly transformative force in telecom, or if it ends up being something that achieves reasonable facsimile status in that it effected change, but ultimately not within scope of the original vision.

On the transformative potential side, there are these two points:

Point 1: SDN-based initiatives such as OpenFlow are being driven by the carrier in response to real-world requirements. This is an important factor given that it highlights a view that the status quo approach of "pseudo" open tools and software applications is too costly, too inflexible and ultimately no longer sustainable.

Point 2: SDN brings some much-needed simplicity to an overly complex world. Given the impact of moving applications to the cloud and the requirement to introduce policy control for application access and security, networks are going to increase in complexity on many levels, and any approach that has the potential to minimize or reduce that complexity is highly desirable. Support of a distributed control plane model is no longer a viable approach.

On the less-than-meets-the-eye side, we have:

Point 1: It's difficult to implement a concept. Despite all the recent activity in the various industry forums, SDN is still largely in the first phase of industry adoption. As a result, it's really difficult to assess where it will be in even the next few years once the real product development work starts to take place. Further complicating the process is that even though OpenFlow, an SDN protocol implementation, clearly has some early market momentum, other approaches exist and more could emerge. For example, in our report we analyze how the IETF's Path Computation Element (PCE) specification may be a more practical approach to take for carrier SDN optical deployments. The question then becomes whether SDN can achieve meaningful cost savings and programmability openness if a number of protocols/specifications that follow a similar methodology are adopted on a global basis.

Point 2: We still don't know what vendors really think about SDN. This is a difficult question and ultimately depends on vendor competitive standing and market momentum. Therefore, vendors will have to tread very carefully as they define their SDN strategies to protect market share while also appearing as aligned to the spirit of SDN and not simply integrating SDN associated buzzwords like programmability into marketing campaigns. This dilemma is further complicated by uncertainty of how licensing of control plane clients associated with approaches such as OpenFlow will be priced.

A lot remains to be sorted out regarding SDN, including potential for success as a game-changer for telecom networks. Over the next 12 months, a number of critical developments both at a vendor and forum level will provide a much better picture of SDN's ultimate impact on the telecom industry.

— Jim Hodges, Senior Analyst, Heavy Reading


For more information about Heavy Reading's SDN & the Future of the Telecom Ecosystem, please contact:




(1)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
obaut
50%
50%
obaut,
User Rank: Light Beer
12/5/2012 | 5:23:42 PM
re: SDN: Start Making Sense





In networks, the events to be controlled, ie packet traffic flows and/or collective bursts caused by individual flows have their natural granularity of in the range of just a kilobit or less, which at eg 10Gbps are through in 10^3/10^10 = 0.1microseconds or less.


Control plane separated from data plane will take seconds to react, ie will unavoidably be thousands of times too slow to be effective.


One could compare the effectiveness of SDN control planes to a thermostat that checks the room temperature twice a year, in January and July, and delivers the adjustment based on the last measurement in six months: it will turn up the heat in July, and the freezing AC in January (assuming the room in question is located in the latitudes of North America).


The software model that worked for managing other software abstractions such as VMs just is ill-suited for handling hardware-level events, such as the more sporadic packetized network communications between VMs at the data plane, and it is even worse for the even less predictable collective packet traffic volume variations between access points of service provider (or enterprise wide area) networks.


Now, we could discuss what the solution is, but that's a bit different topic than SDN.

More Blogs from Heavy Lifting Analyst Notes
Three strategic initiatives &ndash Novitas, Stratus and Sentio &ndash have powered Colt's move into on-demand services.
Next week at BCE, we'll look at some of the ways machine learning and AI will help operators make the customer experience better, without driving up the cost of network operations.
The Aussie operator is trying to reinvent itself, and its network-as-a-service is a vital piece of the transformation puzzle.
Every network operators business case is different but what remains true is that combining the disciplines needed to get value from data requires a very collaborative organization.
Universal customer premises equipment (uCPE) systems are growing rapidly, paving the way for new opportunities, lowered costs and increased performance overall.
Featured Video
From The Founder
Light Reading founder Steve Saunders talks with VMware's Shekar Ayyar, who explains why cloud architectures are becoming more distributed, what that means for workloads, and why telcos can still be significant cloud services players.
Flash Poll
Upcoming Live Events
June 26, 2018, Nice, France
September 12, 2018, Los Angeles, CA
September 24-26, 2018, Westin Westminster, Denver
October 9, 2018, The Westin Times Square, New York
October 17, 2018, Chicago, Illinois
October 23, 2018, Georgia World Congress Centre, Atlanta, GA
November 7-8, 2018, London, United Kingdom
November 8, 2018, The Montcalm by Marble Arch, London
November 15, 2018, The Westin Times Square, New York
December 4-6, 2018, Lisbon, Portugal
All Upcoming Live Events
Hot Topics
What VeloCloud Cost VMware
Phil Harvey, US News Editor, 5/21/2018
Verizon CEO Says LA Is Second 5G City
Dan Jones, Mobile Editor, 5/16/2018
TM Forum Sea-Change Overcomes That Sinking Feeling
Iain Morris, News Editor, 5/17/2018
Animals with Phones
Live Digital Audio

A CSP's digital transformation involves so much more than technology. Crucial – and often most challenging – is the cultural transformation that goes along with it. As Sigma's Chief Technology Officer, Catherine Michel has extensive experience with technology as she leads the company's entire product portfolio and strategy. But she's also no stranger to merging technology and culture, having taken a company — Tribold — from inception to acquisition (by Sigma in 2013), and she continues to advise service providers on how to drive their own transformations. This impressive female leader and vocal advocate for other women in the industry will join Women in Comms for a live radio show to discuss all things digital transformation, including the cultural transformation that goes along with it.

Like Us on Facebook
Twitter Feed