& cplSiteName &

2012 Belonged to SDN & NFV. But Will They Deliver in 2013?

Heavy Lifting Analyst Notes
Heavy Lifting Analyst Notes
Heavy Lifting Analyst Notes
12/19/2012
50%
50%

There's no question (at least in my mind) that software defined network (SDN) and its close cousin, network functions virtualization (NFV), were the big stories of 2012 in telecommunications technology – and not much doubt, either, that they will continue to be the big stories of 2013. But before we can predict they will also be the big success story, an awful lot of detail must be sorted out: We are far from being able to declare definitively that SDN and NFV represent the future of networking technology.

Here are ten challenges that must be resolved if SDN and NFV are to fully realize their huge promise:

  1. How will SDN be integrated with, OSS and BSS systems?
    Presently, this is a big black hole with nothing much in it, but given the legacy that exists in every major telco, it's the biggest unanswered question in the SDN story.

  2. How will the new environment be orchestrated?
    There's a general recognition that an orchestration layer of some kind is required, but will this be accomplished by operators themselves, by vendor proprietary schemes, on the back of open source schemes such as OpenStack or through a new set of standards?

  3. What's the relationship between NFV and SDN?
    Some operators believe that NFV can bring benefits without using SDN (or at least Openflow) – others believe that they are joined at the hip. In 2013, we will see the first fruits of NFV, and with it the beginnings of an answer to this question.

  4. What's the relationship between SDN, NFV and the various telco cloud programs?
    Again, some telcos are trying to ensure that the three developments are coordinated, but though there clearly is a relationship, there's no defined roadmap for how it's constructed.

  5. How far will the ONF be the prime location for SDN development?
    Other initiatives already underway include the IETF's ForCES work, but such is the significance of SDN that we can expect other major industry organizations to get involved too.

  6. Will operators really take the plunge and replace proprietary hardware with generic Ethernet switches and generic industry-standard servers?
    Some already say yes, but when push comes to shove, will the famously conservative network engineering teams agree?

  7. Equally, how will major suppliers respond?
    Despite bold statements from some telcos, few will likely bet the network on start-up vendors, and will likely be dependent on their major suppliers for some time yet. But will those suppliers respond boldly to the new requirements or drag their feet?

  8. How hybrid is hybrid, and for how long?
    The ONF is working on a standard hybrid switch solution, but it's not yet clear what it will look like and whether big established vendors (some of whom already are touting their own hybrid solutions) will play along.

  9. Where will telcos start with NFV and SDN?
    Few, if any, expect a big bang – instead they will likely replace or augment existing networks and functions piece by piece. In its white paper, the NFV group sets out a long list of functions that might lend themselves to virtualization. But where will telcos start, and will they all start in the same place?

  10. Can telcos resolve the many rivalries and tensions among their departments and divisions in a way that enables them to fully realize the benefits?
    This brings us full circle, since the OSS question is right at the heart of this dilemma. Can the gap between IT and networks be bridged in an environment where some functions and divisions may disappear altogether?


It's a long list that raises legitimate questions about the timing of the transition, and it's in the nature of these developments that this list is far from definitive; there are many others.

Making things worse, these questions must be answered in a rapidly evolving environment that may soon include some highly disruptive network service providers using all the principles of SDN to usurp the major telcos and their businesses.

Despite the uncertainties, we should not doubt that the underlying principles of SDN and developments associated are truly revolutionary, and represent perhaps the most exciting potential change in telecommunications technology since IP hit telcos big-time in the mid-1990s. If SDN really delivers, we may find ourselves reversing John Gage's famous 1984 aphorism that the network is the computer; instead, we may see the computer (aka the server, aka the data center ...) becoming the network. As one SDN revolutionary put it in conversation, "Our aim is to make the network disappear."

The stakes could hardly be higher, and we will likely see big fortunes made and big companies lost in the coming transition. Heavy Reading has been following all the key developments closely, and has already published its initial thinking in Multicore Processors Drive the Software-Defined Network: A Heavy Reading Competitive Analysis. And to kick off 2013, we plan a special webinar in early January in which the team will further elaborate its views on SDN, NFV and their potential impact. Look out for that invitation, and in the mean time, a happy New Year from all at Heavy Reading!

— Graham Finnie, Chief Analyst, Heavy Reading


(5)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
tmmarvel
50%
50%
tmmarvel,
User Rank: Light Beer
1/5/2013 | 2:03:27 AM
re: 2012 Belonged to SDN & NFV. But Will They Deliver in 2013?


I guess the here key point is "Will operators really take the plunge and replace proprietary hardware with generic Ethernet switches and generic industry-standard servers?"


A network service provider's major revenues come from 1) Layer 1/0 Private line - (Eth over)SDH/OTN/WDM/fiber, 2) Layer 2 MPLS(-TP) VPN/WAN and 3) Layer 3 IP transit service contracts. There are specialized technologies for providing these contract services. (Generic) Ethernet switches do not allow implementing any of these contract types, whether or not 'software defined' via industry-standard servers.


Where the push comes to shove is when having to answer the question: Does SDN enable providing the network contracts per above with same or better quality, but at substantially lower cost (and why would that fundamentally be the case)?

djthiede
50%
50%
djthiede,
User Rank: Light Beer
1/5/2013 | 2:03:23 AM
re: 2012 Belonged to SDN & NFV. But Will They Deliver in 2013?


shouldn't large enterprises be consider as early adapters?


looks like native generic switches are more suitable for such enterprises than to telcos, isn't it?


looks like telcos are very conservative and it will require some time and more evidences before they will adopt such a revolution.


 

DCITDave
50%
50%
DCITDave,
User Rank: Light Beer
1/5/2013 | 2:03:22 AM
re: 2012 Belonged to SDN & NFV. But Will They Deliver in 2013?


Great way to sum up the issue. You have specialized gear for a reason -- there is big money associated with providing services in a predictable, redundant way with lots of reporting. Can you do this with generic hardware, better software integration and SDN as a controller? Yes or, at least, I bet we're pretty close.


If I were an SP, I don't know if I'd be the first in the pool. Specialized hardware has its place. My smartphone takes a good snapshot but my DSLR is the BEST camera. 


This reminds me of how much guff carrier Ethernet got vs. TDM in its early years. Eventually the specialized service cost too much to provide vs. the commodity service that could do all the same things as TDM.


How close are we to seeing that same thing with SDN in some applications?

tmmarvel
50%
50%
tmmarvel,
User Rank: Light Beer
1/5/2013 | 2:03:20 AM
re: 2012 Belonged to SDN & NFV. But Will They Deliver in 2013?


There is an important undercurrent here however: the need for reduced latencies and jitter (as well as, for certain applications, rock solid reliability and security).


The delay variations (jitter) increase unavoidably when doing packet level switching at each node, as in case of Ethernet. Moreover, when mixing different customers/users' traffic at packet level on same shared switches, there are the everything affects everything problems hurting QoS and causing security concerns. For these reasons, circuit switching T/WDM (either in form of SDH, OTN or direct WDM) has to be used instead of packet switching (eg Ethernet VLANs etc) for latency/jitter critical applications such as networking for high frequency trading and for high quality streaming media transport.


Importantly, the need for latency/jitter minimization is becoming essential for increasingly wide range of applications, eg streaming big data realtime analytics/intelligence and interactive multimedia applications. Multi-user shared packet switching is not a good network technology for these (high revenue) applications.


That is not to say that there could not be 'commodity' hardware that can be managed via vendor-neutral open standard interface from the SP's NMS servers. However that commodity hardware needs to have advanced L1/0 capabilities underneath the packet switched protocol layers to be performance-wise competitive vs vendor-specific NMS+equipment technologies.


In particular I could see a market opportunity for (SDN-managed) pure play L2 MPLS-TP LSR product integrated with L1/0 VPN capabilities. Naturally such technology can be offered also via the IaaS model, ie, contract networks as a service.

spc_vancem
50%
50%
spc_vancem,
User Rank: Light Beer
1/5/2013 | 2:03:16 AM
re: 2012 Belonged to SDN & NFV. But Will They Deliver in 2013?


Using Ethernet it is fully possible to deliver predictable latency with end-to-end jitter in the ns range. OTN and SDH is not required for this purpose.


1) Gigabit Ethernet services can be aggregated/de-aggregated into 10 Gigabit Ethernet pipes with low latency and jitter below 400 ns.


2)  Traffic can be added from gigabit client interfaces onto a 10 Gigabit Ethernet pipe, still with low predictable latency and jitter below 100 ns on the 10G.


Look to TransPacket for more details: www.transpacket.com


Hence, SDN will not inhibit services with OTN and SDH (circuit switched) type of requirement.

More Blogs from Heavy Lifting Analyst Notes
NICs have evolved many times, and the smart NIC is the next step, offering a programmable resource that can be configured to provide additional CPU offload functions for different applications.
Operators are applying artificial intelligence and machine learning technologies to leverage the power of their new programmable, software-based networks.
This year's show evinced healthy interest in effectively using data and analytics to run telecom businesses better, but how well are operators actually doing with it?
FTTx rollouts need a more automated process for collecting and analyzing test results, and analytics could provide the answer.
Driven by web-scale Internet companies, three key trends – disaggregation in terminals, open line systems and 100G+ transponders – are reshaping the DCI market.
Featured Video
From The Founder
Light Reading is spending much of this year digging into the details of how automation technology will impact the comms market, but let's take a moment to also look at how automation is set to overturn the current world order by the middle of the century.
Flash Poll
Upcoming Live Events
November 30, 2017, The Westin Times Square
December 5-7, 2017, The Intercontinental Prague
March 20-22, 2018, Denver Marriott Tech Center
May 14-17, 2018, Austin Convention Center
All Upcoming Live Events
Infographics
SmartNICs aren't just about achieving scale. They also have a major impact in reducing CAPEX and OPEX requirements.
Hot Topics
Juniper's New Contrail VP Hails From Google
Craig Matsumoto, Editor-in-Chief, Light Reading, 11/15/2017
Eurobites: Telefónica Reckons Plastic Is Fantastic for FTTH
Paul Rainford, Assistant Editor, Europe, 11/15/2017
AT&T's Lurie Leaps to Synchronoss as New CEO
Dan Jones, Mobile Editor, 11/17/2017
Animals with Phones
Live Digital Audio

Understanding the full experience of women in technology requires starting at the collegiate level (or sooner) and studying the technologies women are involved with, company cultures they're part of and personal experiences of individuals.

During this WiC radio show, we will talk with Nicole Engelbert, the director of Research & Analysis for Ovum Technology and a 23-year telecom industry veteran, about her experiences and perspectives on women in tech. Engelbert covers infrastructure, applications and industries for Ovum, but she is also involved in the research firm's higher education team and has helped colleges and universities globally leverage technology as a strategy for improving recruitment, retention and graduation performance.

She will share her unique insight into the collegiate level, where women pursuing engineering and STEM-related degrees is dwindling. Engelbert will also reveal new, original Ovum research on the topics of artificial intelligence, the Internet of Things, security and augmented reality, as well as discuss what each of those technologies might mean for women in our field. As always, we'll also leave plenty of time to answer all your questions live on the air and chat board.

Like Us on Facebook
Twitter Feed
Partner Perspectives - content from our sponsors
The Mobile Broadband Road Ahead
By Kevin Taylor, for Huawei
All Partner Perspectives