Addressing the elephant in the tubes
By stretch | Thursday, June 11, 2009 at 3:43 p.m. UTC
Cisco recently released a report predicting massive growth in Internet video traffic over the next several years. Arguably the most provocative prediction made therein states that "global IP traffic will quintuple from 2008 to 2013."
The Internet, to borrow a line from Douglas Adams, is big. Really big. So an estimated 400% growth in its traffic in just five years is more than mild conjecture. The report goes on to assert that "Internet video is now approximately one-third of all consumer Internet traffic, not including the amount of video exchanged through P2P file sharing," which seems unlikely. Then again, when you consider the sheer infeasibility of accurately arriving at these numbers in the first place, any credibility of the report instantly deteriorates into that of a marketing brochure anyway.
Which isn't to say huge growth shouldn't be expected and welcomed; indeed, most experts agree that this Internet thing is here to stay. However, reports like this one, for all their abstract arithmetic, fail to address the ominous creep of artificial throughput limitations quietly (and not so quietly) being implemented by Internet service providers around the world.
Bandwidth caps have long been a burden of life in countries such as Australia and the UK, one which is slowly spreading to broadband subscribers across North America. Last year, Comcast announced an official 250 GB transfer limit amid much controversy regarding their definition of "acceptable use." Several Canadian ISPs have recently had their bandwidth throttling practices called into question. And of course there's the utter nonsense undertaken by Time Warner.
With such miserly tendencies becoming the norm, Internet growth is sure to fall far short of Cisco's predictions. The question now is whether pressure from network vendors and content providers will be enough to reverse the course undertaken by so many ISPs. There is no denying that the technology to support such increases in throughput is available, but can service providers be convinced to adopt it?
UPDATE: BT leadership apparently doesn't understand how the Internet works. But we're to believe that the business strategies at work here are top notch and residential transfer caps are simply a fact of life? Right.
About the Author
Jeremy Stretch is a network engineer living in the Raleigh-Durham, North Carolina area. He is known for his blog and cheat sheets here at Packet Life. You can reach him by email or follow him on Twitter.
Posted in Opinion
June 11, 2009 at 4:13 p.m. UTC
It also seems that the reason ISPs are capping, is because they don't want to reduce profit by increasing Capex & Opex brought on by upgrading the network to handle the load induced by the orginal (uncapped) business model they created.
It's funny in a way there hasn't been a lot of upgrade in the original telco network either (core upgrade yes, but the grid stays the same).
June 11, 2009 at 4:28 p.m. UTC
Doesn't quintuple the IP traffic = 500% more? :)
June 11, 2009 at 4:32 p.m. UTC
@Tom: Somehow, when writing this I knew someone would bring that up. Quintupling a value is a 400% increase, as doubling a value is a 100% increase; you always start with 100%.
June 11, 2009 at 6:29 p.m. UTC
It's a tough call for the ISPs... They can spend millons to upgrade routing and switching, only to do it again in a few years when the new stuff is obsoleted all on their own dime, or they can start to build in limits, and start collecting fees to help offset the costs. Ultimately they are beholden to shareholders on this, not users. I don't like it, I like to pay a small fee monthly for my 5MB pipe, but it's basic capitalism. They need to make a profit for their efforts or they go bankrupt.
June 11, 2009 at 10:20 p.m. UTC
You guys are about to get shafted with Australian style internet, we've had caps for a long long time on our internet. Hopefully it's fixed when they roll out the national fibre network
June 11, 2009 at 10:22 p.m. UTC
In terms of Australian download limits, it is an unfortunate combination of economic and geographical factors.
As most of the content on the internet is outside of Australia, we have had to run massive undersea pipelines to Singapore / Hawaii / Japan. Obviously someone has to pay for these, and it's going to be passed down to consumers.
In terms of internet growth, from what I have read a lot of the infrastructure we have now came out of the dot com boom, leaving us with extremely cheap connectivity after the bust.
Global IP traffic may attempt to quintuple in the next 4 years, but I'd hazard a guess and say this is going to lead to congestion rather than content.
June 12, 2009 at 9:51 a.m. UTC
Jeremy, stop the crusade, it's not going anywhere :) The basic facts are:
Constant growth requires constant upgrades which are hard to do @ fixed low monthly price.
Minority of the users consume majority of the bandwidth. Why should everyone else pay for the upgrades? Contrary to some beliefs, socialism is not a great thing (believe me, I've been there :).
There is no free lunch :)
June 12, 2009 at 1:02 p.m. UTC
@Ivan: While I respect your view, I have to disagree. There are far too many examples of ISPs (and municipalities) which have adopted modern infrastructures to support exponential growth in broadband speeds to accept any excuses masking an inadequate business plan. Verizon's FiOS is a great example of this in the US, and they'll be laughing all the way to the bank in five years.
June 12, 2009 at 1:04 p.m. UTC
My real issue with this is that it can't be fixed by simple free-market competition. Unsatisfied with the state of things, a new small business can't run thousands of miles of its own fiber to compete directly with big ISPs and telecoms who have been in business for decades and certainly aren't looking for friendly competition. (Ironically, new ISPs would consistently sprout up overnight back in the days of dial-up.) The market would be much more open to competition if the last-mile media were instead municipality-owned and subscribers had control over the ISP which they were connected to, but that's a topic in itself.
June 12, 2009 at 1:24 p.m. UTC
I attened the last Network General user conference before they were bought out by Netscout, attending they had a fellow by the name of Jarvis who I believe was the CMO of Forrester Research. He basically said then the same thing that Cisco is saying now.
So for me personally while it is a huge statement for Cisco to make, its not the first person I have heard say it. Which just makes me think its more likly to be that way then not.
June 12, 2009 at 1:50 p.m. UTC
@stretch: that's how it is in a lot of cities in Sweden actually. In my city, the city fiber network is owned and maintained by the municipal. Then all ISPs has to connect to this network make their service available to us. Via a web interface i choose which ISP i want to connect with and can change it every three months.
June 12, 2009 at 9:45 p.m. UTC
Interesting to hear from Cisco. They must take longer to research things before going public. Network World had an article in May about it too. Capacity and IPv6 concerns were brought up. http://www.networkworld.com/columnists/2009/050609-johnson.html?page=1
Wouldn't WiMax become an option for startup ISPs? They'll still need to tie into the 'net somewhere, but I'm sure their product can be cheaper (avoiding tariffs).
June 13, 2009 at 9:25 a.m. UTC
I'm sorry Jeremy, but your lack of experience is really showing here. It's always easy to stand on a soapbox and point fingers when you aren't the one making the business decisions/designing or operating these networks/dealing with the customers.
The reality is that it's just not that profitable to run an ISP. It's a zero-sum game with today's rapidly advancing technology and the subsequent demands that will only increase. You can't build out the infrastructure fast enough, and despite what people may think, there's no such thing as infinite scale. What choice do the ISPs have? Even if Time Warner had the time and money to forklift every bit of their infrastructure from their edge to their backbone with hot-off-the-shelf gear, by the time they were able to finish they'd find themselves again needing to double their capacity (and that's being modest).
June 13, 2009 at 9:26 a.m. UTC
The other commonly forgotten-about factor in all of this is oversubscription. How can a network engineer possibly justify the cost of 100 10G ports when he can't know for sure if all of them will be utilized and in what ways? In the end, the users that use the most bandwidth are the ones complaining about the caps/advocating municipal last-mile control, while the users that don't use as much bandwidth are also complaining because they're getting shafted by the users that do. The technology changes too fast, and the companies that run these networks have to spend the money on the infrastructure and squeeze every last ounce out of it before they can even think about upgrades.
June 13, 2009 at 9:26 a.m. UTC
Assuming that TW and Comcast have gobs of capacity just sitting around doing nothing and that their goal is to nickel and dime their customers for extra cash is naive. If it were profitable to run an ISP, and the cost of buying and operating the infrastructure was inversely proportional to bandwidth demands, we, the end-users, wouldn't be seeing the caps/tiered pricing models, would we?
June 13, 2009 at 5:13 p.m. UTC
@jon: And what experience do you have managing an ISP that gives you more authority on the matter?
For all your rhetoric, you fail to address the numerous examples of service providers who have done precisely what you describe as unprofitable. The difference lies simply in initiative; some providers are proactive, others are reactive. Stop making excuses for poor business strategy.
June 13, 2009 at 6:56 p.m. UTC
continuing my response - sorry for the multi-part comment
Last-mile control is largely irrelevant anyway. Assuming all last-mile access wasn't controlled by a telco, does it really make a difference if your local city or town delivers a cable to your curb if your bits still have to get off of their backbone and onto a T2/T1 (most likely the same one that would have delivered the cable to your curb in the first place) 80% of the time? Not only that, but what would be cheaper? Paying TW a little extra or the taxes that would be imposed by operating municipal last-mile access? Oh, and you can forget about SLAs and timely upgrades too. As I'm sure you know, nothing is done quickly in the government.
As Ivan pointed out, there's no such thing as a free lunch.
June 13, 2009 at 7:12 p.m. UTC
Baiting me into a pissing contest/dismissing my points as rhetoric? Smells like a lot of deflection. Who are you claiming has good business strategy? Verizon? Correct me if I'm wrong, but in one post you're saying Verizon's FiOS is a good example of a broadband product, while in your post immediately following you badmouth those "big ISPs and telecoms that have been in business for decades". Are you completely missing the fact that Verizon is the 2nd largest telco in the U.S. next to AT&T? Make up your mind.
Do you really think you can make an apples-to-apples comparison between Verizon (who's main market share isn't even in home broadband access) and players like Time Warner and Comcast? Verizon operates a Tier-1 network. Wake up.
Please, provide the list of broadband service providers that have pulled off the magical feat of keeping their heads above water with infrastructure costs and have large amounts of ample capacity, and are also making money hand over fist.
June 13, 2009 at 7:19 p.m. UTC
And, no, I don't work for TW or Comcast, and I'm just as annoyed about the bandwidth caps as the next guy. :)
June 13, 2009 at 7:20 p.m. UTC
@jon: You haven't answered question. What experience do you have in managing an ISP?
If you intend to condescend me on my own site you had damned well better have something better than your above wiggling to back it up.
June 13, 2009 at 7:34 p.m. UTC
Chill out. Take your "Defensive Hat" off. Feel free to interpret "lack of experience" as "lack of exposure" or whatever you want. My intent was not to condescend, but to try to get you to think more objectively before you spout off with how evil/terrible/etc these broadband providers are. The point is, no matter who you are, if you haven't designed/built/ran/financially managed a service provider network, then you automatically a pass to have an opinion that's unfounded.
As a consumer of said services? You absolutely have every right to be pissed off. I'm pissed off, a lot of people are pissed off. But at a minimum, take a step back and think about it before you start evangelizing, that's all.
June 14, 2009 at 4:39 p.m. UTC
So what you really meant by "your lack of experience is really showing here" was "I don't have the slightest bit more experience with the matter than you." Must have been a typo, right?
Read the comments here for the inconsistencies between your claims and what is available in some parts of the world.
June 14, 2009 at 11:15 p.m. UTC
Hi Stretch! Thanks for the blog. I wanted to chime in about the internet availability in other parts of the world... You simply cannot compare Japan to the U.S. Japan is physically smaller than the state of California and is 10x more densely populated than the U.S. The infrastructure footprint is incredibly smaller and easier to upgrade. Perhaps a mass migration from the burbs to large cities could help future proof our internet connections?
June 15, 2009 at 8:54 a.m. UTC
Is there anything in history that would shed light on the internet's future.
In the past computers were extremely expensive. As they became more and more common the price dropped and just about everyone can afford one these days.
Won't the same happen with network infrastructure?
June 15, 2009 at 4:47 p.m. UTC
I'm curious why you seem to have such a stuck bit here. I appreciate that if ISPs offer certain throughput rates, they should honor those rates. On the flip side, though, what's wrong with paying for consumption? Look at the "old" days of CIR, burst rate, etc, and those billing structures as an example.
Once upon a time it was easy to calculate over subscription rates since "fat pipes" were T1s and customer based services were ISDN. With the advent of broadband to the home and the huge to-the-home pipes offered in some areas, there has been "keeping up with the Jonses" going on in the networking world (my neighbor has 12 meg service - 768k is for l00z4z), as well as horrible over subscription ratios being commonplace.
Restructuring the cost of internet service is inevitable - ISP is not like the old days of Bell or ATT long distance service. ISPs are generally not making money hand over fist and laughing at the consumer. It's a tight business to be in, IME.
June 15, 2009 at 4:47 p.m. UTC
Maybe a better way for me to phrase this is what "right" does a consumer have to expect more for less on a scale that shifts several orders of magnitude every few years? My MIL who sends 2 emails a week and reads about the same doesn't need 12 meg pipes and would never hit a cap measured in GiB. My brother who games 16+ hours a day and generally has at least half a dozen torrents going at any one time may need more pipe and might hit a cap regularly. Is there some reason that billing them accordingly isn't "fair?"
Please understand, I'm not trying to be inflammatory here, I'm honestly shocked when I hear people with major networking chops rail against usage caps.
June 15, 2009 at 6:58 p.m. UTC
Why would you expect a cheap hard disk you buy today to have more capacity than an expensive one purchased five years ago? You have a very reasonable expectation that open competition will drive innovation, and this is indeed the case, but not in the world of residential Internet service. In the US, providers typically hold a monopoly over their distribution medium. With no or few competitors who offer such a vital service, what motive is there for an ISP to enhance its offering?
Also, the corollary of the "more for less" defense asserts that Internet access speeds should never increase without a proportionate increase in cost. Unless you honestly believe 100 Mbps to the home should cost in excess of $400/mo, it's time to drop that argument.
June 15, 2009 at 7:31 p.m. UTC
If you provide a service to your customers, and your customers demands grow to exceed your capability, then why would you not expand your operation to satisfy it?
If you own a theater and you are constantly selling out of tickets for a show, wouldn't you consider building a second theater so you can sell more tickets? Or maybe you own a store, or operate a power company, or build widgets... it's just common sense business.
It's not as if ISPs thought their network would satisfy demand forever; they must have known they would need to upgrade eventually, so why is there resistance to do it?
I suspect at least part of the answer is, as stretch said in the above comment, lack of competition. Customers aren't going anywhere else if they only have one broadband option (as I do), so sticking them with an aging network with ridiculous cap rules to keep it running smoothly is a viable "solution".
June 15, 2009 at 7:44 p.m. UTC
The offering is being enhanced, though - in general, throughput is increasing, latency is decreasing, and reliability of "Five Nines" is commonplace.
Hard drives: bit density on your platters doesn't affect anyone else, and look at the PC market in general - razor thin margins is a generous description. You're also comparing a commodity item to a shared medium with real MRC.
If you want to compare it to something, how about water? You can get yourself a personal high pressure line, and it will deliver water at an astonishing rate, but if you actually open the spigot and let loose with a swimming pool's worth of water, or just run your sprinklers 24x7, you pay for the water you're consumed. Yes, I realize that water is a finite resource and bits-per-second isn't, so how about electricity instead?
I have no idea what FE to the home should cost, but a $400 price tag wouldn't shock me - DS3 costs well over that if you include mileage, local loop, etc.
June 15, 2009 at 8:22 p.m. UTC
@Snarkout: "...reliability of 'Five Nines' is commonplace."
I think a number of people would disagree with you here. Remember that these are shared residential services we're discussing. Additionally, the trend toward transfer caps ensures that growth in throughput speeds will quickly taper off.
The analogy you provide isn't applicable as the transfer of data is not akin to the consumption of a physical resource.
June 16, 2009 at 5:50 a.m. UTC
Nope, that wasn't what I meant, and nope that wasn't a typo. I know my experience level. I also know your experience level (as you've published it in more than one place on your website). That's all you're going to get from me on the matter. Did I say you were stupid? Not at all. Publishing multiple technical articles with the level of depth and detail that you provide takes quite a bit of dedication, and that is indeed admirable. Due to the previous, you've also developed a set of readers, which is also fantastic. However (and maybe it's as a result, but it's what annoys me the most), you've made the big mistake of believing that it's okay to publish opinion articles as if though you were an authority on the subject. Not a good move.
June 16, 2009 at 5:50 a.m. UTC
However, as you point out, this is your site, so nobody is going to stop you from doing so, and as long as that happens, someone that does have two legs to stand on is going to come by and challenge you, and may succeed at making you look silly. You'd better be prepared for that (and not with weak rebuttals and anecdotal evidence). BTW, the comments on your other blog entry are again, purely anecdotal, so they won't suffice. Sorry. You also haven't addressed any of my previous points.
June 16, 2009 at 9:14 a.m. UTC
Hey Jon, see where it says "posted by stretch in Opinion" up there? Not sure how to make that declaration more obvious. Maybe a blink tag? Although, incidentally, it seems I'm at least as much as an authority on the subject as you, so why don't you hop off that high horse? I at least put my name next to what I write.
I'm not going to keep going back and forth with you on this. I've made my points and they stand firmly, although I have learned the issue here isn't as apparent as I assumed it was. Maybe I'll write a full paper expressing my views at some point. Until then my time is dedicated to other endeavors.
June 16, 2009 at 9:57 a.m. UTC
1. New infrastructure needs to be added by ISP's : YES 2. Bandwidth caps will apply : Unfortunately YES 3. Is it good and acceptable : NO!!! 4. Why is it not acceptable : Because we don't want to! Customer's never worry about how the provider gives it, they demand more. 4. When will it go away : Near future for sure, when competition and pressure from the Inet society gets too much, or when they decide to cut profits.( which i doubt will happen ). 5. Am i an expert : Hell No! I'm a rookie. I guess that's why i see this as black and white, as it actually is IMHO. 6. Do i wanna argue with a "more experienced guy telling me i'm good but not good enough " : No thanks!
However Kudos Stretch! You've kept up the high content standards you've set and maintained throught! Cheers.
June 16, 2009 at 1:09 p.m. UTC
"Hey Jon, see where it says "posted by stretch in Opinion" up there? Not sure how to make that declaration more obvious."
Yes. You don't need to make it more obvious, you need to write your opinion articles more objectively. Try to be non-biased. Make your points. Don't assume your readers will agree with you. Learn to take criticism.
"Although, incidentally, it seems I'm at least as much as an authority on the subject as you"
No, you aren't.
"I at least put my name next to what I write."
Uh... what? Okay, tell you what, if I want to post a comment to one of your blogs in the future, I'll just make up a name. How does that sound?
June 16, 2009 at 1:17 p.m. UTC
"4. When will it go away : Near future for sure, when competition and pressure from the Inet society gets too much, or when they decide to cut profits.( which i doubt will happen )."
Please stop assuming that the providers' profit margins are high. Chances are the caps are never going to go away, unless of course we all want them to go out of business.
"5. Am i an expert : Hell No! I'm a rookie. I guess that's why i see this as black and white, as it actually is IMHO."
Humility is good, but the issue isn't as simple as you think it is.
"6. Do i wanna argue with a "more experienced guy telling me i'm good but not good enough " : No thanks!"
Stick to writing technical articles then, they're either accurate or they're not. Stay away from opinion pieces until you become that expert that you currently claim you're not. Or at least write them very carefully.
June 16, 2009 at 4:45 p.m. UTC
It seems to me that ISPs are the victims of their own short-sightedness. In trying to land grab customers they dropped prices so ridiculously low that the margin is near zero and at no point did any of them have the intelligence to realise that the network was going to reach saturation point in the extremely near future.
Mind you when the network is run by BT it's hardly surprising this has happened.
June 16, 2009 at 9:19 p.m. UTC
@Jon: At this point you're just trolling, so I'm going to end it. However, feel free to write up and send me a link to your own article or paper, something attributable (e.g. tied to your professional identity as my own work is), and I'll link it here.
July 13, 2009 at 6:22 p.m. UTC
To over-simplify it:
Any IT rookie knows that while hardware capabilities increase, prices more or less stay the same or even decrease (I can buy a much more capable router today for the same price I could have bought it for 5 years ago). Any business undergrad is familiar with equipment depreciation it's effects on tax liability, etc. So it's only natural that as old equipment is deprecated and removed from service, capacity increases for a stable inflation indexed price. And I'm not even getting into the economies of scale advantage. So just by upgrading your equipment every few years (along with smart engineering) pretty much takes care of the "ever increasing" bandwidth requirements. Problems arise when one entity shows monopolistic behavior and try to use their scale to push out competition and increase prices.
Wished you'd stop the trollers earlier.
[MBA, MCSA, CCNP, CISSP]
September 29, 2009 at 6:55 a.m. UTC
According to me it is completely depending your ISP. They are spending a thousands of dollars for routing and switching but my problem regarding internet is not solved. I have been complaint for that issue but no solution is work for my slow internet connection and they are charged a 30 dollars per week from me for wireless connection.
January 20, 2011 at 1:32 p.m. UTC
We might know what's the problem beyond this slow connection because I am also struggling in this one.
We should hire/expert skilled technician on this so that we will not keep spending 30 bucks a week.