…the real question for today’s broadband networks is not whether they need to be managed, but rather how.

We couldn’t agree more, since I’ve expressed that same sentiment once or twice or thrice. Network managers know that networks need to be managed. Cerf even explained why:

Network capacity (bits per second or data rate) is a limiting factor in all communications networks. Users cannot send traffic faster than the amount of network capacity available to them. But when users’ aggregate demand exceeds the available capacity of the network, network operators naturally seek to manage the traffic loads… The end result is the potential for traffic congestion, leading to service delays and even outages for consumers.

Cerf then goes on to discuss various methods, such as transmission rate caps, low latency prioritization and bandwidth constraints, but they’re all based on that phrase: “…not whether [networks] need to be managed, but rather how.” [Emphasis added.] I take great interest in these remarks, because I’ve been following his arguments over the last couple years.

For example, back on June 13 of 2006, he appeared on Public Radio’s The Kojo Nnamdi Show. You can find that episode online; if you skip ahead to about 23 minutes in, you can hear Cerf speak of net neutrality for a few minutes. At that time, he argued that innovation will be stifled and users will not be able to freely access content unless the Internet is kept “open and neutral.” He said that the government may need to provide protection. Two years later, it’s unclear what innovation was stifled.

By October of 2007, Cerf was speaking of other concerns, such as at his address at the WebbyConnect conference, which I attended. He said in his speech that “net neutrality” as a phrase has been distorted and that he would clarify what Google was asking for.

• It’s okay to charge for higher capacity.
• It’s okay to address denial of service attacks, viruses and so on.
• It’s okay to provide low latency services.
• As long as all of these practices are done in a non-discriminatory fashion.

Even at this point, Cerf was arguing that network management was necessary, but he had his opinion about what methods ought to be used.  That approach is much better than comments (like this one) that argue that the “management” argument is a scare tactic. There are those who would claim we should just build a bigger pipe. But then you read this quote from a Singapore ISP executive: “Even buying more bandwidth will not work since stuff like BitTorrent is designed to gobble whatever extra bandwidth we buy.”

Cerf specifically mentions that conversations with Comcast engineers have led him to a better understanding of the underlying motivation and rationale for that company’s network management decisions. As we often see, when people examine the capacity constraints we face and the unique challenges of running cable systems, it can contribute to the conversation.

Now that Cerf has confirmed that network management is a fact and a necessity, we can begin to have the broader dialogue about the network management that needs to take place. Cerf argues the “how” of network management is the important piece of the puzzle, but I would suggest the “who” is even more critical – as in, “Who decides what network management practices are reasonable?”

Cerf argues that government should. I think that is the wrong approach. I think it makes more sense for engineers and companies to make those decisions, not government bureaucrats. Those decisions should be based on what methods of network management might be most efficient and which ones would provide the best experience for the largest number of their customers. They should not be based on the dreams and schemes of politicians.

I recently heard one of these people describe network congestion as akin to a mail truck being full. This kind of talk makes my head hurt.

I shudder to think of the regulations that would be dreamed up by the US Department of Network Management.

• Richard Bennet of Broadband Politics: A sad day for the Internet
• Randolph J. May of the Free State Foundation: Hard Cases Make Bad Law: On Regulatory Bits and Torrents–Part II
• Barbara Esbin of The Progress & Freedom Foundation: Legally and Procedurally Suspect and Misguided as a Matter of Policy
• Scott Cleland of The Precursor Blog: Takeaways from FCC Decision on Reasonable Network Management
• Cynthia Brumfield at IP Democracy: FCC’s Copps: Apply Net Neutrality Principles to Wireless Carriers
• Stacey Higginbotham at GigaOM: FCC Punts on Network Neutrality
• Matthew Lasar at Ars Technica: Reactions to FCC’s Comcast decision come fast and furious
• Drew Clark at Broadband Census: FCC Hammers Comcast For Deception and Unreasonable Internet Practices

UPDATE: Will Richmond at VideoNuze: The FCC’s Comcast Sanction: More Problems, Fewer Solutions Ahead

In response, the NCTA  has released the following statement from Kyle McSlarrow, our President & CEO:

“One need look no further than today’s FCC decision for proof that engineering challenges on the Internet should be solved by engineers, not government officials. In second-guessing reasonable network management techniques (with no notice or guidelines in place) that benefit the overwhelming number of broadband subscribers in America, the FCC has inexplicably elevated the interests of a few bandwidth hogs over everyone else.”

As we further digest the decision, we will be following the reaction from others in the telecom policy sphere and will comment further and share what others are  saying.

As I wrote just last week, it’s critical that we can all agree with the principle that “some kind of network management is necessary to ensure a quality experience for our customers.” Once we get past that concept, we can discuss and debate what’s the best way to achieve the goal of a quality Internet experience, but we can hopefully also agree that the government is not the best body to make these decisions.

In this morning’s Washington Post, FCC Commissioner Robert M. McDowell poses the question: Who Should Solve This Internet Crisis? He outlines past network challenges and describes how “engineers, academics, software developers, Web infrastructure builders and others” came together to find solutions. He then answers his own question.

The Internet has flourished because it has operated under the principle that engineers, not politicians or bureaucrats, should solve engineering problems.

P2P apps present particular challenges for network managers, as McDowell acknowledges, and just building bigger pipes doesn’t fix the problem. That’s not to say that this challenge (and others) can’t be addressed. McDowell points out that we need to avoid creating a bigger problem.

Our Internet economy is the strongest in the world. It got that way not by government fiat but because interested parties worked together toward a common goal. As a worldwide network of networks, the Internet is the ultimate “wiki” environment — one that we all share, build, pay for and shape. Millions endeavor each day to keep it open and free. Since its early days as a government creation, it has migrated away from government regulation.

If we choose regulation over collaboration, we will be setting a precedent by thrusting politicians and bureaucrats into engineering decisions.

It’s a good idea to take a look at the whole thing, because it illustrates an important point.  I hope we made the argument sufficiently in our post that some kind of network management is necessary to ensure a quality experience for our customers.  This online discussion illustrates that achieving this is a complex issue. Almost any decision requires you to balance pros and cons. It’s complicated and it’s not clear what the correct path is, which then probably requires a period of some experimentation.

Given all of that, why would you want to put a government agency in charge of deciding what particular method of management should be used?  Or worse, have it decide that no methods of management can be used at all?

“Once more unto the breach, dear friends, once more,” cried William Shakespeare’s Henry V in the play so titled. “Or close the wall up with our English dead!” Perhaps in said spirit did the National Cable and Television [sic] Association (NCTA) veep Michael Schooler and Insight CEO Michael Willner march up to the eighth floor of the Federal Communications Commission on the ninth of July to plead the cause of ISP “network management”…

Taken in conjunction with yesterday’s post on DSL Reports, it sounds like we painted a portrait of Biblical apocalypse. NCTA’s own Michael Schooler and Insight’s Michael Willner supposedly warned of “the impending destruction of the Internet by P2P users.” Or else we said “that the Internet would all but collapse.”

Wow! That sure sounds scary. But since neither Karl Bode nor Matthew Lasar was actually at that meeting, they instead apparently based their accounts on a letter we filed. If you read it for yourself, you find that four points were made.

• Network management is necessary to prevent serious congestion.
• Service for customers would be degraded without such management.
• Network upgrades alone won’t solve problem.
• The government should not pre-determine the tools and technology to be used for network management.

So I ask: Which of these four points are in contention? The DSL Reports post even says “Most techs don’t oppose reasonable network management (booting extreme gluttons, some QOS and prioritization)…” So, we can start by agreeing that reasonable network management is a good thing. Without some kind of management, problems will arise.

Let’s look at service degradation. Was complete congestion claimed? The phrase used is “can cause substantial (and sometimes complete) congestion of the system’s upload capacity.” Let’s emphasize three key words: can, sometimes and upload. This is critical, because peer-to-peer applications are the focus of attention.

This goes to the point about simply upgrading a network. A peer-to-peer application looks for users with the best upload connection. Building a bigger pipe does not eliminate the necessity of network management.

Finally, is the federal government really the best body to judge what network management tools are appropriate? I’m not convinced it is. Nor am I convinced that the answer is a big dumb pipe that treats all bits equally, whether it’s a phone call, streaming video, a P2P download, an e-mail, or a Web page request. And anybody who actually understands how networks work wouldn’t either.

Both of these posts claim that we are crying “Armageddon!” for nefarious reasons. But should nothing be done at all? We want to give our customers the best Internet experience possible, now and in the future, and we need network management to accomplish that goal.