Light Reading

Learning From Mistakes

Jeff Finkelstein

    If you're not failing every now and again, it's a sign you're not doing anything very innovative. – Woody Allen

Or, as I like to look at it, you never learn anything by getting it right. Most great lessons are learned when you are digging yourself out of a hole. A friend is someone who throws you a rope to get out, but a best friend is the person in the hole with you.

Many years ago I worked on a project involving a software system called the Distributed Computing Environment (DCE), which came out of another project I worked on called Project Athena (but that is a story for another blog). DCE was a toolkit for developing client-server applications using remote procedure calls (RPC), location services, time synchronization, authentication services, and a distributed file system (DFS).

While DCE did not gain much acceptance in its initial incarnation, we did learn a lot about how to decompose programs into functional components and have servers in the network run those specific functions. By using the location services, we were able to have servers execute specific sub-programs or functions, based on a variety of selection options (e.g. proximity, spare compute cycles, or other criteria.) DCE was developed in response to dissatisfaction with mainframes and super-minis being costly, hard to manage, and difficult to scale in hardware, as well as power and cooling issues and rackspace requirements.

We find ourselves at a similar, though not quite identical, place in today's cable architecture. We may continue to increase the density of edge devices, just as we did in early 1990s, but is that really the right path to take? We can easily ride the curves of Moore, Koomey, and Denard for some time to increase the density of access equipment. However, as we learned in the days of the mainframe and super-minis, if we fail to learn from our mistakes to prepare adequately for future changes, we are highly likely to make the same poor choices again.

As part of the thought exercise, let's consider our current edge devices, including cable modem termination systems (CMTSs) and optical line terminals (OLTs). Today's DOCSIS edge devices perform many functions, primarily L2 (switching) and L3 (routing). Some functions are what traditionally would be part of an edge router, others are more like an aggregation switch, and many are similar to a media converter.

As each set of actions is scaled, the size, power, and complexities of the edge device will require either more dense and capable silicon or a larger form factor. An OLT traditionally is a L2 device with the L3 functions handled in a broadband remote access server (BRAS). But, with Converged Cable Access Platform (CCAP), we are considering a hybrid that manages both layers in a single device, like a CMTS does today.

So how do we begin thinking about separating the different layers? Which tools are available? What do the layers look like? And, more importantly, why would we do it?

This is where software-defined networking (SDN) and network functions virtualization (NFV) enter the discussion. To get things started, let's begin with a diagram of the higher-level functions performed within a DOCSIS edge device (e.g. a CMTS or CCAP).

Thanks to Harmonic Inc. for allowing me to use this diagram:

Source: Harmonic Inc.
Source: Harmonic Inc.

As you look at this image, you can see the top-level functions that a DOCSIS edge access device performs. As we begin to consider ways to use SDN and NFV, this functional breakdown will be our starting point for decomposing functions.

In my next blog, I will start discussing possible ways to break down the functions into layers and begin to think about separating those layers into a distributed access architecture.

Until then...

— Jeff Finkelstein, Executive Director of Strategic Architecture, Cox Communications

(4)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
User Rank: Light Beer
1/21/2014 | 5:32:16 PM
Re: Performance?
To me both SDN and NFV are not about performance, they are about scalability. I would not expect us to compare them at that level.

Reliability is a great question. In our networks today we have many components where we build in redundancy. They tend to be higher in the network where they aggregate a significant number of customers or if lower in the network it is usually power components that tend to fail or line cards so we can perform in-service upgrades without impacting customers.

Once we reduce a serving group size to 125 or so HHP, how much redundancy do we need? I would say this needs study and analysis by those that track reliability to a component and eco-system level. I am very interested in hearing what these engineers have to say about it.
User Rank: Light Beer
1/21/2014 | 5:26:51 PM
Re: Which functions first?
Ray, I agree that there will be similarities, but there is also a basic difference in how we both approach the network. The access devices in the telco domain tend to have much fewer customers than in the cable space. As a result separating out L2 from L3 is in some ways simpler. In cable with more functionality at the edge we have an inherently more complex box which makes separating functions more difficult.

If I was going to make a guess, I would think that the functions that are already defined in standards (PCE - Path Computation Element for one) would be the first we will see. As new hardware is developed over time there is an opportunity to rethink how we build our cable access networks. The amount of compute cycles needed remains, but where we place it is an important choice to make.


sam masud
sam masud,
User Rank: Light Sabre
1/21/2014 | 10:13:33 AM


Along the lines of Ray's question, I'm curious whether you have any concerns at this point about the performance and/or reliability of NFV?
User Rank: Blogger
1/20/2014 | 2:09:00 PM
Which functions first?

Which function(s) do you think will be subject to the NFV treatment first in cable networks?

Curious to see how it relates to the similar debate in telco networks.

More Blogs from Column
Adding virtualized local content delivery capabilities to broadband networks would bring content closer to consumers and boost OTT quality.
As OTT video traffic soars, CSPs can use new probing tools to measure and improve their data customers' quality of experience more effectively.
The edge router's days are numbered, argues Scott Sumner.
Don't panic – the platform isn't burning but is evolving quickly through open source, writes OPNFV director Heather Kirksey.
Cisco has once again set the bar and challenged the New IP networking sector by opening its doors and welcoming independent testing, something more vendors should be doing.
From The Founder
Steve Saunders provides an overview of white box networking and introduces a new "slim line" version of the OSI 7-layer model.
Flash Poll
Live Streaming Video
CLOUD / MANAGED SERVICES: Prepping Ethernet for the Cloud
Moderator: Ray LeMaistre Panelists: Jeremy Bye, Leonard Sheahan
Between the CEOs
Meet the Architect Behind LinkedIn

12|1|15   |   10:46   |   (0) comments

Steve Saunders speaks with Saikrishna Kotha, who leads LinkedIn's data center infrastructure architecture strategy. With over 400 million users, it's Kotha's job to deliver agility and speed, but also act as a business strategist to develop new ways that LinkedIn can monetize its user base. According to Kotha, LinkedIn runs on a mix of white box and vendor ...
Telecom Innovators Video Showcase
Service Gateway Virtual Edition

12|1|15   |   8.29   |   (1) comment

Allot's Jay Klein talks to Steve Saunders about its new Service Gateway Virtual Edition.
Wagner’s Ring
The Business Case for Open Source

11|30|15   |     |   (3) comments

Open source helps companies enhance infrastructure, undermine competitors' advantage and teach collaboration. But there are problems.
Between the CEOs
Centec on Ethernet Switching

11|26|15   |   09:58   |   (0) comments

Centec CEO James Sun talks to Steve Saunders about Ethernet switching and the white box revolution.
LRTV Custom TV
Delivering Service Agility in the Virtualization Era

11|25|15   |   5.41   |   (0) comments

Interview with Massimo Fatato, WW OSS Business Lead, Hewlett Packard Enterprise.
Wagner’s Ring
How Might Open Source Fail?

11|24|15   |     |   (10) comments

Open source, SDN, and NFV are looking inevitable – but performance, standards proliferation and regulatory capture could derail the movement.
LRTV Custom TV
NFV Lifecycle Orchestration – a Fresh Vision for Telco

11|23|15   |   6.40   |   (0) comments

Simon Osborne, CTO Comptel, and Heavy Reading's Caroline Chappell reveal the business impacts of new SDN and NFV, and what the term service orchestration actually means. Together they define Lifecycle Service Orchestration and how the virtualized future will look for telecoms operators.
Between the CEOs
Cisco's Virtual Role in Saudi

11|20|15   |   12:15   |   (2) comments

Light Reading founder and CEO Steve Saunders talks with Zayan Sadek, Regional Manager at Cisco Systems, about the competitive communications services market and advance of virtualization in Saudi Arabia.
LRTV Huawei Video Resource Center
Huawei Leads With Kubernetes for Cloud PaaS

11|19|15   |   08:26   |   (0) comments

Huawei is looking to Kubernetes as a key tool for building robust open source technologies for customers and partners, said Ying Xiong, chief architect of cloud platform at Huawei, in an interview with Light Reading West Coast Bureau Chief Mitch Wagner at the recent Kubecon conference.
Women in Comms Introduction Videos
WiC in London: The Highlight Reel

11|19|15   |   5:33   |   (1) comment

NetCracker's Mervat El Dabae headlines an inspiring morning in London with help from leading women from Vodafone, TalkTalk, Hyperoptics and Ciena.
LRTV Documentaries
Why Saudi's So Hot for New Tech

11|19|15   |   05:07   |   (0) comments

Light Reading's Steve Saunders reports from Saudi Arabia, a hyper-competitive market desperate to embrace the next generation of communications technologies and services.
LRTV Custom TV
Why Data Models Deliver More Value Than Information Models

11|19|15   |   5.08   |   (0) comments

Stefan Vallin argues that more automation is needed to manage end-to-end services and the hybrid networks they run on, and that data models are key to achieving this.
Hot Topics
Samsung: No Sale of Wireless Unit
Dan Jones, Mobile Editor, 11/25/2015
Facebook in Africa: Beauty or Beast?
Iain Morris, News Editor, 11/27/2015
Cable Gives Thanks for Business Services
Alan Breznick, Cable/Video Practice Leader, 11/27/2015
Hong Kong Moments
Ray Le Maistre, Editor-in-chief, 11/26/2015
How Amazon TV Could Own Christmas
Mari Silbey, Senior Editor, Cable/Video, 11/30/2015
Like Us on Facebook
Twitter Feed
December 15, 2015
Virtualizing Cable Services
Webinar Archive
BETWEEN THE CEOs - Executive Interviews
Steve Saunders speaks with Saikrishna Kotha, who leads LinkedIn's data center infrastructure architecture strategy. With over 400 million users, it's Kotha's job to deliver agility and speed, but also act as a business strategist to develop new ways that LinkedIn can monetize its user base. According to Kotha, LinkedIn runs on a mix of white box and vendor hardware but has the DNA of a web-scale company.
Centec CEO James Sun talks to Steve Saunders about Ethernet switching and the white box revolution.
Cats with Phones
Live Digital Audio

Broadband speeds are ramping up across Europe as the continent, at its own pace, follows North America towards a gigabit society. But there are many steps to take on the road to gigabit broadband availability and a number of technology options that can meet the various requirements of Europe’s high-speed fixed broadband network operators. During this radio show we will look at some of the catalysts for broadband network investments and examine the menu of technology options on offer, including vectoring and for copper plant evolution and the various deployment possibilities for FTTH/B.