Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
The Internet

"Exaflood" Disaster Appears Unlikely 72

I Don't Believe in Imaginary Property writes "By now, we've all heard of the 'coming exaflood' that will drown the ISPs in data and smite the wicked P2P users. Fortunately, the 'exaflood' is unlikely to be a disaster. Internet traffic growth is falling year-over-year, and there's plenty of core bandwidth — now handling about an exabyte a month in fact — but the last mile is still slow. So there's a reason that Comcast & co. are worried about losing to P2P, but the Internet itself isn't likely to suffer a meltdown any time soon. And there's plenty of data to counter anyone who says otherwise."
This discussion has been archived. No new comments can be posted.

"Exaflood" Disaster Appears Unlikely

Comments Filter:
  • Exafloods? Listen, buds
    We got the cure:
    Lots o' suds
    Burma Shave [slashdot.org]
  • by TripMaster Monkey ( 862126 ) on Tuesday April 15, 2008 @08:19AM (#23075824)
    You can bet that, despite this hard data, the RIAA and MPAA will continue to spread this FUD as much as possible...anything to salvage their fatally broken business models.
    • by risinganger ( 586395 ) on Tuesday April 15, 2008 @08:39AM (#23075978)
      Of course it won't. They flat out lied about everything else in their claims so why stop now. Hell even they have admitted certain numbers were fictional [insidehighered.com] but that doesn't seem to stop them continuing to use them.
    • by eebra82 ( 907996 ) on Tuesday April 15, 2008 @09:01AM (#23076174) Homepage
      There is very little they can prove, which is why they throw spaghetti on the wall. Of course they want to portray the accused as evil as possible. Most lawyers do this to dehumanize and humiliate the suspect, because when you have people judging people, this type of psychology actually helps.
    • by unlametheweak ( 1102159 ) on Tuesday April 15, 2008 @09:09AM (#23076256)
      The opposite of FUD, whatever you may call it, from the headline article on Slashdot:

      Internet traffic growth is falling year-over-year
      Note quite, but as stated in the referenced article, the rate of growth is falling. Yes there are concerns with increased growth, especially since much of the usage and growth in usage is not the typical text based Internet of the 1990's, but of the multimedia and P2P type growth of the 20th Century. It makes sense that network capacity should keep up with this growth. This seems to be the concern with most people I believe. From the article; "But from 2002-2007, the growth rate has dropped, and it now hovers at 50 to 60 percent a year." This isn't shabby growth by any means, think of compounding effects of this over the long term, and P2P growth is at only 100% a year; again, if you think of it as money compounding one could get rich very quickly.
      • by Anonymous Coward

        The "growth" is the "growth rate" is the "rate of growth". They all mean the same thing, specifically, the change in size over time. To say that growth is falling is exactly the same as saying the rate of growth is falling: the rate of change-in-size-over-time has changed over time.

        Yes, you're correct in pointing out that it's the growth which is getting smaller, not the size, but you're incorrect in your assertion that the summary is misleading.

        • Yes, you're correct in pointing out that it's the growth which is getting smaller, not the size, but you're incorrect in your assertion that the summary is misleading.
          Perhaps not misleading; or maybe I'm just tired. I at first got the impression that this phrase was downplaying the "rate of growth" aspect.
        • by mwlewis ( 794711 )
          No, I'd says he's right on. I'd interpret "growth" to mean an absolute number of bytes by which it's growing (i.e., "** PB"). I guess you could interpret that either way, but I think "growth rate" would be more appropriate. If you measured a tree, and it was 10ft tall, then next month it was 11ft, and the following month, 12ft 1in tall, it grew more in the second month, but the rate of growth was slower.
        • Isn't that just as misleading? I think you actually mean "the positive deltas in the growth rate". Let's try to be a little more precise next time, hmmm?
          • by yuna49 ( 905461 )
            Most processes like Internet growth have long-term curves that look like the logistic function [wmueller.com]. Growth starts slowly, then ramps up quickly, but eventually the rate of growth slows. Sometimes this phenomenon represents a mathematical reality; rates of "penetration" of technologies into a population are the most common example. (In the US, it looks like HDTV's are in the rapid growth phase, but that will slow as the demand for the TVs is sated.)

            Internet traffic growth obviously isn't constrained the same
            • Re: (Score:2, Insightful)

              by besalope ( 1186101 )
              To put this is business terms, the Internet has its own Product Life Cycle (PLC) curve. And if the rate of growth is starting to decline, then we're nearing the end of the growth stage and crossing over into the maturity stage. Sadly, this move to maturity won't affect the 13 y.o.s that constantly plague the online community.
    • The RIAA/MPAA aren't the ones complaining about being drowned in data, it's the ISPs, like Comcast.

      Still, I'm not sure I see the point of this study. Everybody knows the last mile is the main problem. Most of the Internet is in the last mile! Just as your body contains something like 100,000 kilometers of capillaries, but only a few meters of arteries. So saying the problem is "just" the last mile isn't saying much.

      • by Daengbo ( 523424 )
        I get 100Mb/s to my door in a rural area with no complaints about my 100GB/month bandwidth. No last mile problem here. Yay! At least South Korea is good for something.
      • by Shakrai ( 717556 ) *

        Just as your body contains something like 100,000 kilometers of capillaries, but only a few meters of arteries

        This is totally offtopic, but I was going to call bullshit on your 100,000km figure as it seemed pretty excessive. Imagine my surprise when I found this [buddyproject.org], which actually claims that an adult may have upwards of 100,000 miles (160,934km) of blood vessels in their body.

        Learn something new every day I suppose.

  • by Silver Sloth ( 770927 ) on Tuesday April 15, 2008 @08:21AM (#23075844)
    Ok, so my experience is rural UK based but for me the last mile is what stops me using t'internet for realtime video downloads. Sure I do plenty of bittorrent downloads where I can go away and leave them to cook but realtime still sucks and it's all about the slow response over the last mile.

    Now, when they fix that... but maybe by then they'll have increased the backbone as well.
    • by Chrisq ( 894406 ) on Tuesday April 15, 2008 @08:32AM (#23075928)
      That sounds pretty likely. If I could get real time video over the internet I would watch all sorts of things instead of terrestrial tv. Having to start a download and wait for a couple of hours for a half-hour program limits my use and therefore the bandwidth I use. In this way the last mile bottleneck reduces my usage of the core.
    • Re: (Score:3, Insightful)

      by Idbar ( 1034346 )
      In my opinion, that's what's holding the growth of traffic.

      The last mile in many places is still limited. However, I think that as the time passes, people is more aware of the different services that they can find and use online. I remember not too long ago, we stressed the networks, sending 40MB emails with videos with my friends, as we start finding sites like youtube, we stopped abusing the email system, replicating mails. There's still people sending huge powerpoint presentations through mail, but cha
  • Wait for H5N1 (Score:2, Interesting)

    by R2.0 ( 532027 )
    Almost every single company out there has plans for a flu epidemic that consist of 1 line - "work at home on the internet". So they modeled it and - shocker - the system collapsed PDQ. It wasn't switches exploding, but everything slowing to a crawl so that it would be damned near impossible to actually get work done.
    • Re:Wait for H5N1 (Score:5, Informative)

      by garett_spencley ( 193892 ) on Tuesday April 15, 2008 @08:35AM (#23075944) Journal
      How exactly do you model something like that ? Did they model just a specific class of worker (sysadmin / programmer etc.) or did they assume that everyone in their model required x amount of data transfer ?

      Because most of the jobs that I can think of that could be performed at home on a computer don't require a lot of Internet access. Maybe transferring one or two files from the office network but not any kind of constant data transfer back / forth.

      Then you factor in that with so many people at home they'll probably be spending more time slacking off / surfing the net. But people do that at work anyway (I'm a webmaster and I see traffic spikes Monday morning after a weekend slowdown which suggests that people spend most of their time surfing the net from work) so I'm just wondering how you even begin to go about modeling something like that ?
      • Re: (Score:3, Insightful)

        by sexybomber ( 740588 )
        ... with so many people at home they'll probably be spending more time slacking off / surfing the net.

        Wait, refresh my memory. Is H5N1 the flu strain with the ~70% kill rate, or is that something else? The severity of the illness will definitely affect the amount of network traffic produced from people working from home. 'Cause if 100Ks/millions of people contract a really serious flu strain, I doubt they'll be doing much of anything beyond, you know, screaming, moaning, and dying.
        • Re: (Score:3, Insightful)

          by jamesh ( 87723 )

          'Cause if 100Ks/millions of people contract a really serious flu strain, I doubt they'll be doing much of anything beyond, you know, screaming, moaning, and dying.

          I think the point is that everyone will be stuck at home because they are afraid they _might_ catch something.

          Anyway, what will kill the internet is everyone blogging about the tiniest sniffle they get if there is ever a flu scare, and then recording video's of themselves coughing and submitting it to video sites and asking "does this sound like t

      • Because most of the jobs that I can think of that could be performed at home on a computer don't require a lot of Internet access. Maybe transferring one or two files from the office network but not any kind of constant data transfer back / forth. Then you factor in that with so many people at home they'll probably be spending more time slacking off / surfing the net. But people do that at work anyway (I'm a webmaster and I see traffic spikes Monday morning after a weekend slowdown which suggests that people spend most of their time surfing the net from work) so I'm just wondering how you even begin to go about modeling something like that ?

        Working from home - for many this is going to require a VPN connection, especially for people not accustomed to working remotely. Thats going to require some talk back and forth to keep the tunnel open. Workers accustomed to being on the LAN probably have their mail clients configured in a chatty manner - I know I have different Notes settings for office and remote use, and my office setting is very chatty and noticeably impacts my network responsiveness when out of office. Security software, Active Di

    • by Chrisq ( 894406 )
      Also pretty pointless. Unless you spend years in isolation the chances are you will be exposed to the virus.
      • Yeah, it'll be the delivery person for my first home food delivery that'll pass on the disease. I'll have to leave a note on the door - "knock, deliver, and go away".

        Prepare for H5N1 now! Grow your own veg. Just think of the money you'd have saved in 2007 ... 2006 ... 2005 ... whenever this end-of-humanity bird flu was meant to strike.
  • Yes but... (Score:5, Funny)

    by crohan ( 1028032 ) on Tuesday April 15, 2008 @08:55AM (#23076110)
    And there's plenty of data to counter anyone who says otherwise.

    So long as everyone does not access these copious amounts of data simultaneously ;-)
  • I'm actually interested to hear if anyone has coherent suggestions as to how musicians might make money without royalties from albums - or if anyone feels like the free download of music doesn't impact the royalties from these albums significantly, or somesuch.

    You always read/hear people saying that the RIAA has a "fatally flawed business model", and I think that's true. It does seem like there's going to be very little in the means of defense of intellectual property in the future. However, instead of la
    • Here's a hint. You're starting from the flawed assumption that MUSICIANS make much money at all from album royalties.

      Albums should be given away as promotional material for performances (if ever pressed at all instead of just delivered electronically to begin with). Artists themselves already make most of their money from selling merchandise at performances.

      Labels make money peddling plastic discs. Artists pretty much get the shaft.
      • Re: (Score:1, Interesting)

        by Anonymous Coward
        I know that 1 example does not a business model make, but check out Jonathan Coulton [jonathancoulton.com]. He's never touched a major label, but he does fantastic stuff (of both the geeky and less-geeky varieties), and I hear he makes a pretty good living for himself.
      • Re: (Score:3, Informative)

        Or how about:
        - Downloads are free
        - Sharing is free
        - Physical media (CD's, LP's, DVD's) are sold at a reasonable price that is low BUT also gives profit
        - Concerts
        - Eliminate record companies, or cut them down HEAVILY

        I do realize this is utopia, as record companies are not ready to cease existence, but if we all share our music, we might just kill them.
    • What happens when scholars can't sell books because as soon as it's published, it's on the interwebs for free?

      Then the universities will have to publish the books, and help the students save thousands of dollars every year at the same time. Without a monopoly-supported, profit-driven market for schoolbooks, effort can be spent on more useful things than writing 10 books with the same content and release new editions every other year to prevent students from buying books second hand.

    • They could do what they do now, which is get a day job like the rest of us.

      You appear to be under the mistaken impression that record companies ever, ever pay royalties.

    • How do people out there on Slashdot feel about the decline of intellectual property rights and does anyone have any good solutions to the problems that this would hypothetically ... plague those who depend on intellectual property for their survival?

      I would have to suggest some combination of: Patronage, which currently provides us with museums, and libraries, (granted that the 'patrons' in this case are the taxpayers); Collectors, who tend to pay for physical objects anyway, even if the book is already available at the library for 0$, but owning it has considerable attraction anyway to many, and the same can go for recordings, just to a lesser extent; Performance, which doesn't apply to everybody, but the cinema experience will I hope always be worth

      • I would have to suggest some combination of:
        Patronage, which currently provides us with museums, and libraries, (granted that the 'patrons' in this case are the taxpayers);
        Collectors, who tend to pay for physical objects anyway, even if the book is already available at the library for 0$, but owning it has considerable attraction anyway to many, and the same can go for recordings, just to a lesser extent;
        Performance, which doesn't apply to everybody, but the cinema experience will I hope always be wor
    • Simple: sell your music. Why take a small royalty from the label when you can sell your music directly to your customers and keep all the money. You have to pay the bills, but I'm betting you can arrange hosting and an e-commerce site for a low enough cost that you can sell cheap enough to have people buying while still making more actual profit than label royalties would be. The key is to shake loose of the idea that the only way to make money selling music is to collect royalties from a label. Once you do

  • That's because Internet traffic is growing faster than capacity, but also because of the difficulty in upgrading the edges of the network, not just the center (where such upgrades are relatively simple).

    The "internet" can handle it, the problem is with ISPs? Sounds like a good place for competition and innovation to work. If an ISP cannot handle the traffic they sell to their customers then they shouldn't be an ISP.

  • Repeat after me... (Score:3, Insightful)

    by slashname3 ( 739398 ) on Tuesday April 15, 2008 @10:55AM (#23077610)
    How many times in the past have people reported that the imminent failure of the Internet (or USENET) would occur? Get it straight folks, the Internet as we know it may change but it will not collapse or crash or fail. It will adapt as it always has and continue on.

    This is another non-story that should not have been posted.
    • I can understand not RTFAing, and in rare cases the summary, but seriously, the headline? This is an article refuting that claim, providing arguments (and quantitative evidence to support them) against the exaflood scaremongering.

      But perhaps your right - slashdot should only post the outrageous corporate propaganda articles, and ignore the writings that challenge such FUD, leaving each individual to judge BS in isolation.
  • Well, everything is fine when you look at historic trends, but like condo speculators imagining that prices will always go up, looking behind you won't prevent you from falling over the cliff.

    The future of internet has nothing to do with the capacity of the big pipes behind the ISPs, but the cost of it to the ISP. I don't mean that as capacity increases cost reduces, but that as users demand more the ISP must offer more capacity at the same price - users wont accept less (especially if they're promised unli
  • Peak Bandwidth? (Score:2, Interesting)

    by Arngautr ( 745196 )
    A bit analogous to peak oil? (except for the whole, not actually deletable thing...)
    • Except in every way, yes, exactly. Oil is a fixed resource whereas bandwidth is by definition a rate; we need to consume oil at a constant rate simply to keep the system going, but sustained bandwidth is easy. Oil gets harder and harder to extract the more we consume, but we can always add more bandwidth without being penalized by how much we already have.
      • Pending doom and gloom about supply not being able to match demand. There are those who effectively argue that there is no need for concern. Sounds analogous to me...
        • Except that the Evil Corporations (TM) are on opposite sides. Big Oil fights peak oil scares, while AT&T promotes the notion of a bandwidth crunch.
  • Now that traffic shaping & throttling is rolled out across most major ISP's I imagine there's really no impetus to lay down some decent pipes anymore.
  • Don't you love second derivatives?

Life is a healthy respect for mother nature laced with greed.

Working...