Slashdot Mirror


Blackout Shows Net's Fragility

It doesn't come easy wrote to mention a ZDNet article discussing a recent outage between Level 3 Communications and Cogent Communication. A business feud inadvertently highlighted the fragility of the Internet's skeleton. From the article: "In theory, this kind of blackout is precisely the kind of problem the Internet was designed to withstand. The complicated, interlocking nature of networks means that data traffic is supposed to be able to find an alternate route to its destination, even if a critical link is broken. In practice, obscure contract disputes between the big network companies can make all these redundancies moot. At issue is a type of network connection called 'peering.' Most of the biggest network companies, such as AT&T, Sprint and MCI, as well as companies including Cogent and Level 3, strike "peering agreements" in which they agree to establish direct connections between their networks. "

287 comments

  1. The small should pay for the big? by hkmwbz · · Score: 5, Interesting
    As I understand it, these were about the same size and had an agreement, or didn't bother to bill each other. Then suddenly one of them figured out that "hey, we are bigger, so they should pay us!"... And the smaller one cut off the connection because they didn't want to pay since they considered themselves to be as big as their rival.

    What I don't get is why one of them would suddenly want the other to pay up. What's changed now, and why does the smaller company have to pay the big one's bills?

    Am I missing something here?

    --
    Clever signature text goes here.
    1. Re:The small should pay for the big? by Daniel+Boisvert · · Score: 2, Funny

      Am I missing something here?

      Yes. :)

    2. Re:The small should pay for the big? by hkmwbz · · Score: 3, Funny

      Ah, thanks! That certainly made it clearer ;)

      --
      Clever signature text goes here.
    3. Re:The small should pay for the big? by mopslik · · Score: 1

      Am I missing something here?

      Yes. "Greed".

    4. Re:The small should pay for the big? by Daniel+Boisvert · · Score: 5, Informative

      NANOG has been on fire with posts about this issue over the past few days. The following two from Leo Bicknell do a good job of explaining why this sort of thing would happen, why nobody in particular is The Bad Guy[tm], and why this issue has no relevance to the issue of internet resilience in the case of natural or manmade disaster:

      http://www.merit.edu/mail.archives/nanog/msg12302. html
      http://www.merit.edu/mail.archives/nanog/msg12350. html

    5. Re:The small should pay for the big? by Cally · · Score: 4, Informative

      Check the NANOG archive over the last few days for far, far more than you ever wanted to know about "The Art of Peering: The Peering Playbook"... or read the book yourself.

      --
      "None are more hopelessly enslaved than those who falsely believe they are free." -- Goethe
    6. Re:The small should pay for the big? by billcopc · · Score: 1

      Mob rules!

      If the little guy doesn't pay, the big guy whacks the little guy. Welcome to Capitalism 101.

      --
      -Billco, Fnarg.com
    7. Re:The small should pay for the big? by 99BottlesOfBeerInMyF · · Score: 2, Informative

      Am I missing something here?

      I only read about this very briefly, but my understanding is it went beyond that. Just cutting the peering connection is fine and proper and packets then are rerouted through other peers, possibly costing more money, possibly not. Then the internet goes on as before and everyone is happy and the peers involved can negotiate a new link if they want and figure it will save them money by avoiding other routes where they have to pay for traffic.

      My understanding is that in this case they not only cut the link, but they advertised routes to their other peers for traffic from the first peer, which they then maliciously, and probably in breach of those other contracts, filtered out, resulting in failed traffic routing. Basically they intentionally lied (to the routers) and said sure we'll route those and then did not.

      I don't think this highlights the fragility of the internet, so much as the fact that end users usually rely upon a single peer (ISP) and if they can't trust them to not intentionally break traffic they had better find a new ISP.

    8. Re:The small should pay for the big? by Gomker · · Score: 1

      The thing you are missing is the traffic level.
      The peering agreements are, in theory, setup for mutual traffic levels.

      Since one is sending 3 times the amount of data (assuming this is true from article) they think its not fair and should be compensated for the increased amount of traffic.

      Sounds like they could have a case, if the traffic levels are really that skewed.

    9. Re:The small should pay for the big? by Feyr · · Score: 2, Informative

      that's wrong, no one is filtering them. not anymore than they normally would to maintain their network.

      what we are seeing here is a pissing contest between two "tier1". so there literally is no other route the packets can take to reach each other network (contractually speaking, not technically). each of these networks have peering contracts with other companies, not transit. a peer is only used to reach other's network, a transit lets you reach networks beyond the network you are transiting through.

    10. Re:The small should pay for the big? by Skye16 · · Score: 1

      Be this as it may, how is a contractual dispute that disrupts the internet somehow more acceptible than a hurricane or nuclear weapon? End result: I can't see a good chunk of the internet. The cause is completely superfluous to the result. I can't see a good chunk of the internet.

    11. Re:The small should pay for the big? by NanoGator · · Score: 3, Funny

      "NANOG has been on fire with posts about this issue over the past few days."

      WHat?! No! I haven't said a word about it!

      --
      "Derp de derp."
    12. Re:The small should pay for the big? by Anonymous Coward · · Score: 0

      And the smaller one cut off the connection because they didn't want to pay since they considered themselves to be as big as their rival

      Level 3 is the larger tier-1 ISP (according to themselves) and they were saying that they want Cogent to pay for peering with their network. When Cogent replied with "we are the same size and won't pay" ... Layer 3 cutoff Cogent. The larger ISP cutoff the smaller one for refusing to pay. Rediculous ...

    13. Re:The small should pay for the big? by tchuladdiass · · Score: 2, Interesting

      The thing I don't under stand is, say you have two network providers, A and B. If A's customers are sending more data to B's customers, then should A pay B for the route, or should B pay A because it is their customers that are requesting the packets?

    14. Re:The small should pay for the big? by Anonymous Coward · · Score: 0

      Then move to an ISP that doesn't suck, fucktard. Serious, if you "can't see a good chunk of the internet" then your ISP isn't doing their jobs. HINT: real providers have multiple links from more than one provider to guard against outages.

    15. Re:The small should pay for the big? by Anonymous Coward · · Score: 0

      I read about this on GRC.com's news page back on the 5th. He has / had duel T1's with Cognet and servers hosted by Level 3, so he was unable to reach his servers accept through a backup cable modem.

    16. Re:The small should pay for the big? by fafaforza · · Score: 2, Interesting

      Level3's stock price is tanking, they are fighting for survival and the jobs of all their employees, from engineers to secretaries, while Cogent is undercutting the price of bandwidth by a factor of 3 while taking advantage of their peering with not many being able to compete on price, and you call Level3 greedy? C'mon.

    17. Re:The small should pay for the big? by Skye16 · · Score: 0, Offtopic

      Yeah, because everywhere has access to multiple ISPs, fuck tard. I have 1 dialup and 1 cable. That's it. Period.

      Stop with the "it works for me!" bullshit. You just sound like a fucking dick.

    18. Re:The small should pay for the big? by Anonymous Coward · · Score: 0

      duel != dual accept != except; Learn to spell

    19. Re:The small should pay for the big? by Anonymous Coward · · Score: 0

      Dude, you need to relax or your mom's going to take your Internet away. Hey - shouldn't you be in school right now?

    20. Re:The small should pay for the big? by drakaan · · Score: 2, Informative
      That doesn't make any sense. It's not as if there's no other route at all between the two networks. Routing protocols and ICMP unreachables exist to allow traffic to route around trouble like this. Unless the link was deliberately broken and packets unceremoniously dropped, the source for a given connection attempt would see it's packet routed in what appeared to be an excessive manner, but it'd still get from point A to point B.

      If Cogent users can get to Qwest and L3 users can get to Qwest, but cogent users can't talk to L3 users, then cogent and L3 are doing something intentionally bad and screwing everyone on the internet.

      --
      "Murphy was an optimist" - O'Toole's commentary on Murphy's Law
    21. Re:The small should pay for the big? by Skye16 · · Score: 1

      Law #386 of the Intertron: If someone makes a statement alleging you are a minor, they, themselves, are actually minors.

      See psychological studies on insecurity.

    22. Re:The small should pay for the big? by Anonymous Coward · · Score: 0
      duel != dual accept != except; Learn to spell

      Learn to stretch your brain. Dueling T1's is like dueling banjos but quieter.

      DUAL T1s whats AC smokin

    23. Re:The small should pay for the big? by boneshintai · · Score: 1

      It's a business problem, not a technical one. To use your example, if both L3 and Cogent peer with QWest, but neither of them has a transit agreement, QWest is under no obligation to move packets from L3 to Cogent or Cogent to L3. It's a case of minimum contractual conformance dictating the technical details.

    24. Re:The small should pay for the big? by karnal · · Score: 1

      I can't see a good chunk of the internet

      Shit, I've been browsing for about 9 years now, and I still haven't found a good chunk of the internet.

      --
      Karnal
    25. Re:The small should pay for the big? by EvilTwinSkippy · · Score: 1

      That is unless the routers on the Net are still working on the assumption that the shortest route is across the Level 3 / Cogent border. While the internet at large can be quite resilient, a sudden change like this can take days to sort themselves out.

      --
      "Learning is not compulsory... neither is survival."
      --Dr.W.Edwards Deming
    26. Re:The small should pay for the big? by cosmic_0x526179 · · Score: 1
      The thing I don't under stand is, say you have two network providers, A and B. If A's customers are sending more data to B's customers, then should A pay B for the route, or should B pay A because it is their customers that are requesting the packets?

      You really should go read the traffic on NANOG. It goes into great depth to explain the current "lets see who blinks first" situation

      The abridged version is... Cogent and L3 had what is called a peering agreement (this is also called a Settlement Free Interconnect or SFI). Cogent has been low-balling the industry trying to attract business (and possibly undercuting L3). Likely (but unknown) that L3 was handing more of Cogent's originating traffic than Cogent was handling of L3's originated traffic. L3 would like to see Cogent change the relationship from peering (SFI) to Transit (where Cogent would be paying L3 to handle the traffic, mostly because of the imbalance). Cogent would like for L3 to reactivate the peer, so they can go on getting SFI (aka free-transit). Economics and Politics are much involved in this discussion. Go read the 100+ messages on NANOG, there are some great explanations there.

      --
      This msg is brought to you by the letter 'W'.. for Worthless Wuss
    27. Re:The small should pay for the big? by Decibel · · Score: 1

      I don't think the post was about the cause or the result; it was about people ass-u-ming that this was a technical problem. His point is that it's not a technical problem, but a business (political) one.

      The one technical side I wonder about is this: are there good enough tools to allow ISPs to act as transport between NSPs and be compensated (as in: not charged for that traffic)? That's the real reason why the internet of today is 'fragile' is because all the backbone traffic typically goes through a fairly limited number of points, not for technical reasons but for business reasons. If every multi-homed ISP also passed traffic between it's multiple providers the internet would be extremely well connected.

    28. Re:The small should pay for the big? by anticypher · · Score: 1

      Both sides have been playing at stealing the other's customers, it has just gotten very bad in the last six months or so.

      The Cogent offer to existing Level 3 customers requires them to drop their link with Level 3 (and stop paying Level 3 bills). The free year of transit ends if the customer obtains any other redundant transit from another carrier during the year. Its not a very good deal.

      the AC
      I expect both Level 3 and Cogent to be out of business within six months, and then all the other carriers can get back to earning some narrow margins on new contracts as prices rebound to market levels

      --
      Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
    29. Re:The small should pay for the big? by Anonymous Coward · · Score: 0

      Wow. Not many network engineers read slashdot. Or their posts don't get modded up. I just browsed through the comments (and read TFA) and have failed to see an accurate description of the basic problem here. [I'm not a network engineer myself, and I used to work for one of these companies, which is why I'm posting AC.]

      Understand peering: Anybody can peer with anybody; large ISP's peer with each other all the time to save on upstream network traffic(/costs). A peering connection carrys traffic from any of my customers to any of your customers, and vice-versa. Understand transit: When you buy a transit connection from someone, they are saying that they will deliver your traffic to anybody anywhere (you'll _transit_ their network to another network). Most ISPs are just buying transit from some other network (some of which are, again, buying transit from another network).

      Clearly, not everyone in the world can be buying transit. There are a few big boys at the top of the heap that buy transit from no one. These are the "tier 1" networks. For things to work correctly, these tier 1's must all have peering connections to each other. (Don't tell me anything about public peering points, either, please. I'm talking about the real world here.)

      The short story is that Cogent looked at all of the money it was paying for internet service and didn't like it. Then they bought a bankrupt, tier 1 network called NetRail. Then the old tier 1's are grumbling about how Cogent is a big customer one day and a peer the next day. (Nominally, anyway, still lots of contracts, different autonomous networks owned by Cogent, complications galore from the beginning.) They want to keep their club small and shrinking. Anyone they can kick out of tier 1 status must become a customer.

      It's a tough situation -- both sides have a point. Can anyone buy their way into tier 1 status, regardless of their size, and force the others to peer with them? Those connections aren't cheap. Can the tier 1's just gang up on any one tier 1 at any time and decide to cut them off and gain a giant new customer? It's like the paralell parking episode of Seinfeld.

      I'm a conservative and it really pains me to say this, but this is a situation that crys out for regulation. Tier 1 status is a sort of shared monopoly. There's just no way for businesses to not make self interested decisions and, in the case of tier 1's, when they fight we lose. The requirements for becoming a tier 1 (and for staying a tier 1) should be defined by law, and now. Though, a couple of years ago would have been a lot better. This particular fight has been a long time coming (I'm surprised it took this long). Any idiot can see that two businesses fighting over money aren't just going to be able to "work things out".

      So, yeah, there's your solution. But, you may just have to put stamps on your emails one day. :-|

      _ac_

    30. Re:The small should pay for the big? by drakaan · · Score: 1
      I mostly agree. I'd say it's a business problem *causing* a technical one. Is there no way for Cogent and L3 to get the traffic that's no longer moving across the peering points directed to a router that will allow it to reach its destination?

      In your example, you are correct, Qwest is under no obligation to do anything, but L3 and Cogent probably are. I doubt they sold their customers internet access to every internet destination except the ones that are owned by an ISP that they're in a dispute with. It a situation that is incredibly disruptive for certain organizations that use one of those two companies as an ISP, and the reason for the problem is a deliberate choice on the part of the ISP that the ISP *knew* would break some things.

      In the company I work for, we have multiple sites connected via VPN. If one of the other sites hosted an application that users here needed to access, and I purposely broke the VPN connection to that site because of a political dispute with another IT admin, you better believe that it'd be my fault, and that I'd be in deep doo-doo. Why is this situation looked at differently?

      --
      "Murphy was an optimist" - O'Toole's commentary on Murphy's Law
  2. A solution can be... by Yaa+101 · · Score: 1

    A solution can be to make it mandatory that all connecters to any internet exchange point peer to each other. all of them...

    1. Re:A solution can be... by Chang · · Score: 1

      Some of the most critical network interconnects do not happen at "internet exchange points" and the only rules in place are either a contract or just mutual understanding between the companies involved.

    2. Re:A solution can be... by cloudmaster · · Score: 2, Funny

      Be sure to let the UN know about that - this is surely something they'll want to take care of when they take "control of the Internet" away from the US. :)

    3. Re:A solution can be... by ikegami · · Score: 0

      What's an "internet exchange point"? There are no points, just thousands of ISPs. Currently, most have direct connections to only one or two other ISPs. You are suggesting ISPs increase thousandfold the number of connections they have, which is not practical. The internet cannot be fully meshed, not even at the ISP and/or national level.

    4. Re:A solution can be... by Angostura · · Score: 2, Funny

      Wow, that's a nice idea. That'll mean that all I have to do is run a bit of Ethernet into a peering point and I'll get free connections to all the tier ones. Fabulous.

      Oh - hang on, if someone else runs a bit of Ethernet in, do I have to connect to them? Damn.

    5. Re:A solution can be... by Alioth · · Score: 1

      The original poster probably meant an internet exchange like Mae East or Mae West or Linx.

    6. Re:A solution can be... by frost22 · · Score: 2, Informative

      *Sigh*. Why do you spew nonsense if you actually have not even found out how a clue looks like, not to mention ever aquired one ?

      So you claim there are no Internet Exchange Points ?

      pray tell, what is this thing ? Or that one, not to mention the middle one.
      Oh, and what do you think those Guys do for a living ?

      Nobody expects you to be a fucking genius or know everything. But why are some folks constantly touting stupid nonsense instead of keeping their mouths shut and learning something ?

      --
      ...and here I stand, with all my lore, poor fool, no wiser than before.
    7. Re:A solution can be... by Anonymous Coward · · Score: 0
      But why are some folks constantly touting stupid nonsense instead of keeping their mouths shut and learning something ?

      Slashcode was designed to route around this...

  3. make it wireless by LogicallyGenius · · Score: 0

    or use satellite technology

    1. Re:make it wireless by pturing · · Score: 1


      Right, since one can just get dozens of gigabits of satellite bandwidth for free, that solves everything.

  4. No worries by WormholeFiend · · Score: 5, Funny

    The pr0n industry was designed to find alternative routes of delivery in case of Internet outages.

    1. Re:No worries by JPriest · · Score: 3, Funny

      And slashdot runs redundant stories on the same thing in case the first one is lost on the way.

      --
      Saying Java is nice because it works on all OS's is like saying that anal sex is nice because it works on all genders.
    2. Re:No worries by Trogre · · Score: 1

      Yes because as we all know the pornography industry has headed the invention of the computer, the digital watch and the question mark.

      The scary part is, there's probably someone who actually believes this stuff.

      --
      "Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
  5. When did this blackout happen by SolusSD · · Score: 1, Interesting

    yesterday? I swear the entire net came to a crawl yesterday afternoon.

    1. Re:When did this blackout happen by varmittang · · Score: 2, Informative

      I think since Wednesday.

      --
      -----BEGIN PGP SIGNATURE-----
      12345
      -----END PGP SIGNATURE-----
    2. Re:When did this blackout happen by pturing · · Score: 1


      If you're not a roadrunner user, and you don't have people on roadrunner who expect your site to be up all the time, then it's no surprise you didn't notice.

    3. Re:When did this blackout happen by SolusSD · · Score: 1

      i am a roadrunner user. I did notice. did you read my above post?

    4. Re:When did this blackout happen by pturing · · Score: 1

      replied to wrong comment, oops

      I will now go commit seppuku

    5. Re:When did this blackout happen by nihaopaul · · Score: 1

      wow, now you know how it is for us in china. especially via sprintlink, can you say 2000ms of latency? they claim china is over using their bandwidth.

    6. Re:When did this blackout happen by SolusSD · · Score: 1

      wait no!! Dont!!! It was an honest mistake!!! (*&* it.. i'm probably too late. ;)

  6. Didn't notice at all. by Oz0ne · · Score: 1

    I've been hearing about this for days, but haven't noticed any slow links, or down sites... anywhere. I hit lots of major sites a day, minor too. I have a few webservers in various parts of the world... and I didn't have a single blip anywhere.

    What exactly got blacked out? I know the issue, but just never saw it's effect anywhere.

    1. Re:Didn't notice at all. by Ewan · · Score: 1

      It's only a problem if you're on the Level 3 or Cogent backbones, noone else is affected. If you're a level3 user, you can't get to cogent sites, and vice-versa.

    2. Re:Didn't notice at all. by Anonymous Coward · · Score: 0

      What exactly got blacked out? I know the issue, but just never saw it's effect anywhere.

      You would only be affected if you, either directly or indirectly through an ISP, belonged to one of the two feuding networks and attempted to get to certain services provided by the other network. If you don't fit that description, you probably aren't going to see much if any fallout from this.

    3. Re:Didn't notice at all. by lostlogic · · Score: 3, Informative

      You would only notice if you are on one of these two networks. I am personally on UUNet at home and MCI at work, and my server is on SpringLink (via Schlund, who I am not familiar with). As a result, all of my traffic is completely unaffected. Customers on a single-homed connection through Cogent, or through L3 cannot see other single homed customers on the other network. The rest of us don't know the difference. The dumb thing that this article points out is that both Cogent and L3 are refusing to route packets destined for each other through the rest of the internet (probably for fear of fucking up other peering agreements by dumping too much traffic on their other peers). I believe there was a comment in the previous thread about this issue saying that traffic in one direction could be routed, but that even return packets were being null-routed at some point, preventing any type of connection from being established.

      --
      --Brandon
    4. Re:Didn't notice at all. by LurkerXXX · · Score: 2, Insightful
      It's not a blackout. This is a FUD story. There are plenty of alternate routes in place to 'heal' a broken link between the Level3 and Cogent networks. The problem is Level3 is deliberately black-holing traffic to Cogent on the Level3 routers.

      If Level3 didn't want to peer anymore with Cogent. That's understandable, it wasn't an even exchange of traffic anymore. They could have done the right thing and simply stopped the peering. Insted, they have decided to be vendictive and filter any traffic to/from Cogent's IP range, even if the traffic is coming through some other ISPs network that Level3 still has peering or paid relationships with.

      One again, the internet routers are perfectly able to find routes, Level3 is just deliberately trashing the packets before they get there. The Internet isn't 'unstable'. Any ISP can filter packets entering or leaving their network, and Level3 has decided to do so in an bad way. This just means Level3 customers should be pissed. This is nothing for anyone to get their panties in a bunch over except Cogent and Level3 customers, who's ISPs are being dicks.

    5. Re:Didn't notice at all. by bubbaprog · · Score: 1

      I am on a verizon fiberoptic line, routed through level3, and my work/school's ISP is Cogent. The past three days have been a disaster for me. Students can't get a hold of me, I can't do research, and my email inbox is over quota because I haven't been able to check it until arriving on campus this morning. This is a SIGNIFICANT problem for a lot of us. Not to mention all the students on campus who can't access any site hosted by level3 -- which is a good chunk of the internet.

    6. Re:Didn't notice at all. by frost22 · · Score: 1

      The blackholing thing seems to be an unsubstantiated rumuor - I've yet to see any postive confirmation from a credible source (i.e. not some guy on ./ ).

      From what I can see L3 simply has no Cogent routes because nobody advertizes them to L3, and L3 refuses to pay someone for that. Same with Cogent. There is no middle Ground ind this conflict - with peering they were already on middle ground. Now L3 essentially demands payment to route Cogents Traffic while it expects the same traffic to pass Cogents Network for free.

      The traffic imbalance shit is just baloney - if there is much more traffic coming from Cogent that means L3's customers have requestet that traffic. There is no "sender pays" principle in the internet, and on packet level this would be utterly ridculous anyway.

      It all comes down to the old mafia principle - I am stronger than you, so you pay me money.

      --
      ...and here I stand, with all my lore, poor fool, no wiser than before.
    7. Re:Didn't notice at all. by TwinkieStix · · Score: 1

      If this is true, As far as Cogent's hardware is concerned, shouldn't it look like L3 is down? Then isn't BGP set up at Cogent in such a way as to recognize that the routes are "down" at L3 and find another path? I thought that this was a standard feture of routers on the Internet.

    8. Re:Didn't notice at all. by LurkerXXX · · Score: 2, Informative
      I have friends working for other Level3 clients and peers. The packets are getting to Level3. Then they disappear. The routes ARE advertised to them. That's the beauty of good internet routing, it heals around wounds.

      FYI, smaller ISPs pay larger ISPs for bandwidth all the time. The larger ISPs have huge costs. Switches costing hundreds of thousands of dollars, filled with a bunch of cards in it that each cost hundreds of thousands. Lots of them. Lots of fiber and other costs. It gets real easy to have billions invested just in hardware. They offset those costs in part by selling bandwidth to smaller ISPs. That's the way the net works.

      Try telling some small ISP that they should stop paying their upstream provider. That the upstream provider should give them bandwidth free so that the larger ISPs customers can access websites hosted by the smaller ISP. They will tell you you are living in a dream world. That's not the way the net works.

    9. Re:Didn't notice at all. by LurkerXXX · · Score: 2, Interesting

      That's exactly what it looks like. And yes, the routers are set to find another path. The problem is when it finds a new path through some 3rd or 4th ISP to the Level3 network, as soon as the Level3 router sees the packet originated from a Cogent IP address, it null routes it. That's not a problem with fragility of the net, it's Level3 behaving badly. (Note: Cogent should have ponied up money for traffic to a larger provider to avoid this mess in the first place. There are no good guys involved in this.)

    10. Re:Didn't notice at all. by onepoint · · Score: 1

      this will help you confirm your information
      just adjust the settings to 24 hours and you'll see
      http://scoreboard.keynote.com/scoreboard/Main.aspx ?Metric=AvailNtwk

      --
      if you see me, smile and say hello.
    11. Re:Didn't notice at all. by frost22 · · Score: 1

      Nope, that doesn't confirm anything. That just shows (showed - they have resumed peering now) lack of connectivity. That doesn't say anything about underlying causes.

      I specifically addressed the rumour of blackholing - i.e. Supression of Traffic by active and wilful false routing ("driving traffic into a black hole"). I continue to assume - as most experts have mentioned - that by cutting the peering Level3 lost the only advertized route to Cogent and the ASes that are only reachable via cogent.

      They obviously speculated/hoped Cogent would pay someone else who already peers with L3 for transit, thereby advertizing their routes, too. This would have harnmed Cogent while not harming Level3, which probably was their objective in the first place, anyway.

      New announced cutoff day is November 9th, now, btw. We can stay tuned :-)

      --
      ...and here I stand, with all my lore, poor fool, no wiser than before.
    12. Re:Didn't notice at all. by TwinkieStix · · Score: 1

      So, Cogent is dumping traffic onto lower tier providers, and the low tier guys have the right to shut Cogent off for it. Cogent has the right not to do business with L3. Nobody has to do business with anybody in particular, but if a provider can't get packets though, it should eventually go out of business. I can't see how Cogent is being bad at all. But, for L3 to block packets coming from these lower tier providers because they happened to pass through Cogent is anti-competitive to say the least.

  7. Background info by NicolaiBSD · · Score: 3, Funny

    Hey, I've found some interesting background info on this novel story here.

    1. Re:Background info by KDan · · Score: 2, Funny

      I think this source is blocked by the filters the editors use when selecting stories... It's pretty untrustworthy and contains lots of potentially inappropriate material...

      Daniel

      --
      Carpe Diem
  8. Efficiency can be the enemy of robustness by dpilot · · Score: 5, Interesting

    This statement popped up in some of my security readings. It's most "efficient" to have one path between two places, and it's most "efficient" to set up peering agreements to route packets. But these efficient measures can introduce single points of failure.

    On a similar note, that's why there are 13 root DNS servers, and why most of us aren't supposed to use them. The DNS example though, is one where efficiency and robustness agree. It's more efficient, at least in terms of net bandwidth, to use a DNS server closer than the root servers.

    --
    The living have better things to do than to continue hating the dead.
    1. Re:Efficiency can be the enemy of robustness by c_g_hills · · Score: 1

      There's a lot more if you count those that use anycast. In addition, there is the open root server network, which follows icann policy, but hosts its servers predominantly in Europe.

    2. Re:Efficiency can be the enemy of robustness by cheezus_es_lard · · Score: 1

      I guess I don't understand why either provider isn't changing their routing tables, shucking all of (say, in L3's case) the Cogent-destined traffic out to the public internet. Instead of inconveniencing your customers, wouldn't it make more sense to make some temporary routing changes, that might end up costing you a little more in traffic, but will prevent these 'blackouts' of network areas to customers with backbone access from that provider.

      Can someone explain why this isn't as easy as I think it is?

    3. Re:Efficiency can be the enemy of robustness by Soul-Burn666 · · Score: 1

      That is not possible since these are top level providers.
      A standard refers to 2 types of connections:
      Customer-Provider and Peering.

      An ISP X publishes, at most, the routes as follows (they can choose to publish less):
      X publishes to its peers and providers the routes to its customers.
      X publishes to its customers all the routes it recieved.

      This creates a layered model, where in the top you have the top level providers, the internet backbone.

      Due to this, it is impossible to route:
      * down to a customer and then back up to a provider (think \_/ )
      * peer-to-peer-to-provider ( _/ )
      * peer-to-peer-to-peer ( _._ )

      Because of this, it is impossible to route it like you mentioned.

      The top level must be a clique of peers or some networks can not be reached.
      Obviously, every ISP wants to be a provider since that gives money, while peering is "free".

      What usually happens is that the link is disconnected, all the customers are complaining. It's like a game of "chicken", until one of the ISPs breaks and becomes the other's customer.

      Hope this explains the situation.

      --
      ^_^
    4. Re:Efficiency can be the enemy of robustness by PhilHibbs · · Score: 1

      This "public internet" of which you speak. What companies are you referring to? Who pays them? Who runs it?

      If they redirect the traffic that would have gone via each other out through their other connections, then their customers (smaller ISPs) would start complaining that Cogent/L3 are abusing the connections that the smaller ISPs are paying them for.

      At least that's how I understand it, my understanding being based on what I have read over the last 24 hours on this issue.

      Phil Hibbs.

    5. Re:Efficiency can be the enemy of robustness by cheezus_es_lard · · Score: 1

      I guess what I'm questioning is why the L3 folks don't remove the Cogent peering routes they're advertising, and allow their traffic to take other paths through other peers (say, Sprint to WorldCom to Cogent) instead of causing this blackout.
      I understand it's 'corporate chicken' but they could accomplish the same thing without inconveniencing their customers, I would think.

    6. Re:Efficiency can be the enemy of robustness by PhilHibbs · · Score: 2, Insightful

      Sure, they have the connections, but routing extra traffic through those peering links will probably only cascade this problem. The intermediate providers will see a jump in trafic coming through from L3 and Cogent, and they will have to consider how to recoup the costs that that is imposing on them.

      It's a web, and when one strand breaks, it increases the strain on the other strands.

  9. Typical by SilverspurG · · Score: 0, Flamebait
    From the article:
    "I have been pushing for years to have a redundant ISP for our traffic," Bradham said. "But we're a nonprofit. We don't have the money available to do that."
    How much does this guy make every year? Less money spent on extravagant CxO and board member salaries means more money for a redundant ISP... and they could still keep nonprofit status.
    --
    fast as fast can be. you'll never catch me.
    1. Re:Typical by SilverspurG · · Score: 1

      Yes. I got the flamebait because I forgot to check that Bradham is the network engineer.

      Line of logic still holds. The museum has plenty of bigwigs and money makers in the upper levels. Maybe they could chip in sometime.

      --
      fast as fast can be. you'll never catch me.
  10. Call the helpdesk...wait, THEY don't even know! by digitaldc · · Score: 4, Interesting

    http://www.gamergod.com/article_display.cfm?articl e_id=329
    Good article on this situation here

    This situation has adversely affected various users of both companies' services. The inability of Level 3 to handle this situation in a fair and equitable manner to the consumers has alienated many customers and will continue to do so until the current situation is remedied. At what point is it good customer service to discontinue services due to no fault of said consumer base? Market history shows us that the single worse thing a company can do is to arbitrarily allow influences beyond the control of consumers to negatively impact services, determined by consumers to be status quo, without any warning or notification. If left unresolved and unaddressed, the current situation could set dangerous precedents for internet users across the country by allowing service providers to instantly discontinue provided services at the moment they feel that the services they provide are not being adequately compensated for from outside companies.

    On a side note, I was listening to Howard Stern (oh no!) this morning and he said that his Time Warner internet connection at home didn't work. Howard then called a tech guy to come and fix the problem, only for him to call a help desk to figure out what happened. The help desk didn't even know what was wrong. It sounds like Level 3 just pulled the plug and didn't notify ANYONE. Or maybe it was Cogent, the point is nobody outside of that dispute KNEW what was going on.
    This sounds like a good way to alienate your customers and/or ruin your business model. But that is just my opinion.

    --
    He who knows best knows how little he knows. - Thomas Jefferson
    1. Re:Call the helpdesk...wait, THEY don't even know! by peragrin · · Score: 4, Funny

      You want scary, I can show you scary. I emailed Roadrunner saying I would drop them if they couldn't due something.

      I got a semi canned response but it did have some techincal details. It also stated that if you wish to discuss the techincal nature of the problem go to www.ask.slashdot.org With a full link to the other article.

      Yep Roadrunner sent me to slashdot to get more information.

      --
      i thought once I was found, but it was only a dream.
    2. Re:Call the helpdesk...wait, THEY don't even know! by pturing · · Score: 1


      heh, guess he called the wrong helpdesk
      I work support for a hosting company on cogent, and knew about it the moment I arrived at work Wednesday

    3. Re:Call the helpdesk...wait, THEY don't even know! by Anonymous Coward · · Score: 0

      I was listening to Howard Stern (oh no!) this morning and he said that his Time Warner internet connection at home didn't work.

      I had internet peering way before any of these guys, Robin. They're all ripping me off. Hoo hoo! Tell 'em, Fred!

    4. Re:Call the helpdesk...wait, THEY don't even know! by nb+caffeine · · Score: 1

      now THAT is either funny or scary... That the sent you to slashdot or full detials (funny) or scary, if you dont have an account and browse at -1...

      --

      "Something's wrong with you...and I hope we never do meet again." - Deftones When Girls Telephone Boys
    5. Re:Call the helpdesk...wait, THEY don't even know! by Hrodvitnir · · Score: 1

      You know what? I commend them for doing that much.

      Call me masochistic, but it makes me feel better to know that the Road Runner techs read /.

      --
      "There are more important things than stopping terrorism. Upholding the Constitution is one of them." - Ars Forumer.
    6. Re:Call the helpdesk...wait, THEY don't even know! by DigiShaman · · Score: 1

      Yes, we do.

      --
      Life is not for the lazy.
    7. Re:Call the helpdesk...wait, THEY don't even know! by hypervinetest45 · · Score: 0

      What's really scary is that you misspelled a two-letter word. "...I would drop them if they couldn't [do] something."

    8. Re:Call the helpdesk...wait, THEY don't even know! by psyon1 · · Score: 1

      Do you get some feeling of superiority when you correct typing mistakes? The poster got their message to us, and we understood it. That is all that really matters.

    9. Re:Call the helpdesk...wait, THEY don't even know! by Thundersnatch · · Score: 1

      I would hate to see your resume.


      In real life, spelling and grammar do actually matter.


      Of course, Slashdot doesn't exactly qualify as "real life"...

  11. designed to withstand? Says who? by Anonymous Coward · · Score: 1, Insightful

    In theory, this kind of blackout is precisely the kind of problem the Internet was designed to withstand

    Riiight...just like the Internet keeps everyone anonymous.

    1. Re:designed to withstand? Says who? by 'nother+poster · · Score: 2, Insightful

      No, the internet was never designed with anonymity in mind, but it was designed to be a communications network that would not experience systemic problems when individual nodes and connections went down.

    2. Re:designed to withstand? Says who? by Tribbles · · Score: 1

      Yes, Bob.

  12. A New Approach by mysqlrocks · · Score: 2, Interesting

    So, it appears a big part of the Internet traffic is controlled by large companies like Cogent or Level 3. No big surprise. I think this highlights the need for a new approach to connecting people together. I know there's been talk of wireless mesh networks where everybody is both an end point and a router. This would work in populated areas but I'm not sure how well it would work for "long haul" connections which is what the issue is here. Can anybody think of (or know of) any alternatives that gives control and power of the Internet back to the people who use it?

    1. Re:A New Approach by dada21 · · Score: 1

      Laws and mandates will make this worse. Only the mighty dollar, YOUR mighty dollar, will make a difference.

      If you use Cogent or Level3, dump them. Find another provider.

    2. Re:A New Approach by SlayerofGods · · Score: 1

      It's easy to get the power back to the people that use it just set up a nonprofit company and start laying fiber.... no the REAL problem is how to pay for it.
      Some how I'd expect voluntary contributions would fall short.
      An internet tax would work.... as evil as it might sound.

      --

      Technology, the cause of and solution to all of life's problems.
    3. Re:A New Approach by mysqlrocks · · Score: 1

      just set up a nonprofit company and start laying fiber

      I live in Burlington, VT and the city is doing this through an organization called Burlington Telecom. I'm one of the beta testers and we're going to be hooked up with a 5 Mbps symmetrical fiber connection in a couple of weeks. The best part? It's cheaper then Adelphia and I, as a citizen, own the network. Unfortunately, this still doesn't solve the "long haul" problem.

    4. Re:A New Approach by fireboy1919 · · Score: 2, Insightful

      I know there's been talk of wireless mesh networks where everybody is both an end point and a router. This would work in populated areas

      This would work in populated areas in theory. In practice, though, 95% of the bandwidth in any given system gets eaten up by 5% of the users unless there is heavy regulation. Actually, we pretty much need the big internet companies in order to get a particular level of QoS.

      Like I said, all it takes is one in fifty who won't play nice to ruin it for everybody else. I'd be willing to bet that 1 in 50 people is a sociopathic jerk - probably even more. Ultimately, we need something to keep the sociopaths from going nuts. "Power to the people," like the anarchy that is mesh networking won't work.

      --
      Mod me down and I will become more powerful than you can possibly imagine!
    5. Re:A New Approach by mysqlrocks · · Score: 1

      Like I said, all it takes is one in fifty who won't play nice to ruin it for everybody else.

      Very good point. Could there be some sort of system to moderate "good" and "bad" nodes? Hmm, kind of like on slashdot?

    6. Re:A New Approach by fireboy1919 · · Score: 1

      ...and I can't do math. 1 in 20 is 5% :)

      --
      Mod me down and I will become more powerful than you can possibly imagine!
    7. Re:A New Approach by BeBoxer · · Score: 4, Informative

      I know there's been talk of wireless mesh networks where everybody is both an end point and a router. This would work in populated areas but I'm not sure how well it would work for "long haul" connections which is what the issue is here.

      If by "work in populated areas" you mean "slow the network to a crawl" then yes, it would work. Mesh networking is cool stuff, but you aren't going to build a backbone out of it. Wireless is really fast compared to your DSL line or cable modem. But it isn't even in the same ballpark as what you can do on fiber. Backbone links are running at 10Gbps or even 40Gbps. Full duplex, so that is 20Gbps or 80Gbps of "marketing bandwidth". Compared to what, 22Mbps or 54Mbps half-duplex for your wireless? You aren't going to build a comparable backbone out of wireless links running at roughly 1/1000th of the speed. Physics pretty much guarantees that fiber links will always be faster than wireless.

    8. Re:A New Approach by cloudmaster · · Score: 1

      Let the UN take over!

    9. Re:A New Approach by TheRaven64 · · Score: 1
      In a way, you are completely right. Fibre is a lot faster than wireless in a straight line. However, a mesh network (which doesn't have to be 100% wireless, by the way - some segments can be wired) has a lot more more potential paths for traffic. It is possible on a mesh network to route sequential packets via different routes, as long as your router is powerful enough to keep up with huge routing tables that may change frequently.

      A high speed mesh network is an engineering problem, not a theoretical one.

      --
      I am TheRaven on Soylent News
    10. Re:A New Approach by Eli+Gottlieb · · Score: 1

      I think I'll bring up something that was mentioned in "Citywide Fiber Project Challenges and Goals" and then ignored in the political quibbling: It is entirely possible to start a cooperative ISP owned by its customers. Such a company, by mandate of its charter and its Board, will HAVE to provide the best service they can at the best price they can with the best reliability they can. And if they don't, the customers can vote them out.

      All without getting the government involved.

    11. Re:A New Approach by BeBoxer · · Score: 1

      It is possible on a mesh network to route sequential packets via different routes, as long as your router is powerful enough to keep up with huge routing tables that may change frequently.

      Are you going to get a thousand paths? Not likely. Each one of those hops is going to introduce latency as well, which hampers TCP performance. Lots of changes in the routing table? If the topology is too dynamic you're going to lose packets in the various transient loops and dead-ends that occur. And your packets are going to be effectively shuffled when they get to their final destination, greating increasing the load that needs to be done on the TCP stream to put it back together. End result? Poor TCP performance. A working mesh network is an engineering problem. A mesh network with performance comparable to a well engineered high speed fiber backbone? Fantasty.

    12. Re:A New Approach by MikeBabcock · · Score: 1

      If we had fiber to the home, and routers on street corners that were ISP agnostic, we'd be getting somewhere ...

      --
      - Michael T. Babcock (Yes, I blog)
    13. Re:A New Approach by bonehead · · Score: 1

      Would you REALLY want your Internet connectivity to be at the mercy of a /. style moderating system?

      You've seen how well it works here.

    14. Re:A New Approach by sploxx · · Score: 1

      Well, then let's use free space optical wireless mesh networks. Practically no interference, no broad antenna beams etc.

      There are very cheap, homebrew solutions.

      I'm proposing this for YEARS now, sadly finding anyone who is willing to build such a thing (even in a city) is a complicated *social* effort :|

    15. Re:A New Approach by BeBoxer · · Score: 1

      Well, then let's use free space optical wireless mesh networks. Practically no interference, no broad antenna beams etc.

      The Ronja looks like a neat piece of equipment. But it's still running at 1/1000 the speed of a real high speed fiber link. As I said, mesh networking is cool, and you can do some neat things with it. But perform the same as a high speed fiber backbone isn't one of them. You never even really addressed my main points. Having a lot of hops increases latency, wireless or not. Complicated and overly dynamic routing will result in lost packets. Multiplexing traffic over multiple links will cause out of order packets. Replacing 11Mbps 802.11b gear with 10Mbps Ronja gear doesn't change the fundamental problem. Physics favors fiber optic cables over free space optics for all the same reasons it favors cables over wireless except one (FSO and fiber optic run at the same frequencies). Fiber optic cables will still have less noise, less loss and the ability to put a bunch of them into a conduit to increase capacity.

    16. Re:A New Approach by tylernt · · Score: 1

      "Like I said, all it takes is one in fifty who won't play nice to ruin it for everybody else."

      That's what dynamic traffic shaping is for. It's already built in to Linux. One guy using up all the bandwidth? Throttle his connection down to 56kbps for a while.

      --
      DRM 'manages access' in the same way that a prison 'manages freedom'
    17. Re:A New Approach by fireboy1919 · · Score: 1

      You don't get to do that. Nobody's controlling the connection. Everyone is connected together in a giant mesh.

      Sounds like you want someone to have control of things.

      Just like I said.

      --
      Mod me down and I will become more powerful than you can possibly imagine!
    18. Re:A New Approach by tylernt · · Score: 1

      If nobody has any kind of control, the abusers will quickly take over and render the network useless. I can't see it working in real life.

      I'm not saying there should be one central controlling entity, just that each person should have control over their own little node. If Joe Neighbor is routing gigs of crap through my node, I better be able to shut him down and make him route his traffic elsewhere, or else my own node is useless to me. Might as well unplug it.

      --
      DRM 'manages access' in the same way that a prison 'manages freedom'
    19. Re:A New Approach by Geoff-with-a-G · · Score: 1

      And what's more, when you talk about the bandwidth of that fiber, you're talking about the bandwidth inside it. You can lay other fibers right next to it in a conduit. When you talk about the bandwidth in wireless (say, 54 Mbps for 802.11g) you're talking about available bandwidth in that channel.

      The impact of that is, if your cell has a radius of 100 feet, then everyone using channel 6 (802.11 has three "non-overlapping" channels, 1, 6, and 11) in that 100 foot radius shares that 54 Mbps. So if we're talking about running an Internet connection from NYC to Boston, in the best of cases, in which nobody runs wireless except participants in our mesh, the total bandwidth available for everyone going between NYC and Boston becomes 54 Mbps. With fiber, AT&T can run a conduit with 10 pairs of fiber. Verzion can run 10 pairs. Qwest can run 10 pais. Now they put an OC-192 on each pair, and you've got something like 300 Gpbs between NYC and Boston.

      And we haven't even started talking about noise/jamming problems. Wireless backbone is a contradiction in terms.

    20. Re:A New Approach by sploxx · · Score: 1

      Fiber optic cables will still have less noise, less loss and the ability to put a bunch of them into a conduit to increase capacity.
      Probably very true :)

      But also consider that a fiberless link can get (or is, in the case of WiFi) ridiculously cheap compared to a fiber. And with FSO, anyone that sees any other one can form a link. Two cheap devices (with the proper economy of scale) on each end, ready.
      With fiber, you have do invest a lot in digging up the streets, the fibers itself etc.

    21. Re:A New Approach by BeBoxer · · Score: 1

      But also consider that a fiberless link can get (or is, in the case of WiFi) ridiculously cheap compared to a fiber. And with FSO, anyone that sees any other one can form a link. Two cheap devices (with the proper economy of scale) on each end, ready.
      With fiber, you have do invest a lot in digging up the streets, the fibers itself etc.


      Sure, I agree with that. I don't think I ever said a good fiber backbone is cheaper than some sort of mesh using wireless and FSO. Or even in the same ballpark for cost. I can go 10 miles across town with gige LX on fiber optic for just a few hundred buck on either end. Oh, and after spending a half a million dollars trenching fiber. ;-)> But my original point was that it isn't feasable to replace current fiber backbones with wireless (or FSO) and expect to get anywhere near the performance.

    22. Re:A New Approach by Anonymous Coward · · Score: 0
      I think this highlights the need for a new approach to connecting people together. I know there's been talk of wireless mesh networks where everybody is both an end point and a router.

      That could work for modest traffic loads within areas with dense connectivity (several alternate routes everywhere). It would not support heavy downloads for a large number of users, as the available transport capacity between A and B would not be impressive. You need a hundred wireless links in parallell to rival one fibre. Ergo, the wireless mesh would have to latch onto fibre to be useable for anything resembling sustained broadband speed. Fibre links between cities are not priced at consumer level!

  13. Re:Ask Slashdot by Anonymous Coward · · Score: 0

    You answered your own question. It's just not possible without servers in different physical locations. You have to account for more then just network trouble, there's always natural disasters to consider.

    Then again, if it's a hobby project out of your own pocket iut doesn't actually have to be 24/7. It's not like you're signing a contract to guarantee that kind of uptime.

  14. Peering by Neurotoxic666 · · Score: 5, Funny

    At issue is a type of network connection called 'peering.'

    In other news, the RIAA announced they've stopped an extremely large P2P network.

    --
    You are more than the sum of what you consume. Desire is not an occupation.
    1. Re:Peering by dpilot · · Score: 1

      Please, don't give them any ideas!

      While the Internet is central to music sales through Amazon and other such stores, and music downloads, I'm sure the RIAA would shut it down in a heartbeat, if they could.

      --
      The living have better things to do than to continue hating the dead.
  15. PEERING! by b3x · · Score: 0, Offtopic

    The RIAA's lawyers are currently working on documents to stop this new threat to the entertainment industry!

  16. Level 2 Providers by Anonymous Coward · · Score: 0

    This is why you make sure that your ISP/provider peers with mutliple "Tier 1" providers.

    We peer through 3 major providers, and our users didn't see an outage because of it.

  17. -1, Redundant by slimey_limey · · Score: 1

    This was reported two days ago. Don't believe the hype!

  18. Internet can route against natural calamities by anandsr · · Score: 4, Informative

    Internet cannot route when your providers do not want you to communicate.
    Nothing can protect you in this case.
    If on the other hand there was a natural calamity and every one was trying to get you access
    then you would get it. Like it happened during Katarina.
    This is not a natural calamity.

    The best option is to ditch your provider if they are not a monopoly and if they are lobby to your government to create multiple providers.

    1. Re:Internet can route against natural calamities by mjh · · Score: 2, Interesting
      Internet cannot route when your providers do not want you to communicate. Nothing can protect you in this case.
      I agree with the first part, not the second part. What protects you if your providers don't want you to communicate is:
      1. contracts that state that your provider is required to allow you to communicate
      2. competition from other providers
      --
      Key to financial independence: Spend less than you earn. Save and invest the difference. Do it for a long time.
    2. Re:Internet can route against natural calamities by jeanicinq · · Score: 1

      The best option is to ditch your provider...

      That is a good suggestion. The pragmatic case of the two rivals yesterday only lay ground evidence why wireless multipoint networks are in demand today. It is also the same reason why such big landline companies don't want to implement wireless multipoint; they'll lose their "providership" when customers can easily route around the provider. In that case, customers might finally realize that they are the Internet and the big self-claimed providers are really not.

  19. Whole company by FoxDude0486 · · Score: 0, Redundant

    Everyone here at the company I work for was crying their internet was down because the msn homepage and some microsoft page were down and they didn't think to try another site. When your job as IT becomes to tell people to try a different website then the homepage then there's a problem.

  20. Whats up? by Anonymous Coward · · Score: 0

    Didn't notice anything different in the UK. Some of my torrents had less peers than usual. Shrug.

  21. It's dupealicious! by mrpotato · · Score: 3, Funny

    But for easy karma, just go get a +5 comment in the other thread, and repost it here without attribution.

    Not that I would ever do such a thing...

    --

    cheers
    1. Re:It's dupealicious! by Anonymous Coward · · Score: 0

      Funny, I saw the exact same comment in the last dupe, and the one before that, and the one before that, and the one before that....

    2. Re:It's dupealicious! by Anonymous Coward · · Score: 0

      I don't see any other mentions that this is a duplicate story .. I wonder if Slashdot is filter or deleting those dupe comments?? ..I better be careful or I might get blacklisted for making such comments..

  22. It always will be fragile by squoozer · · Score: 4, Insightful

    The Internet will IMVHO always be quite fragile. While the design lends itself to robustness the reality is that there is only money for a few very big connections and therefore a disaster that affects one of these connections is going to cause wide spread outages.

    Take, for instance, the connections running between Europe and America. I bet most of them run in almost exactly the same place on the sea bed because it's the cheapest / shortest path to take. A fairly localized geological disaster (at least in geological terms) could cut all the cables at once; or at least enough to make to difference.

    If we wanted the network to be robust we would need to run cables up over the north pole and round the equator and probably stick in some satelite links as well. There just isn't money for that. People are willing to accept the risk that it might fail in extreme situations.

    FWIW I think the problem is worse on the global scale than the country scale. I imagine most developed countries probably have enough redundancy in their own country. It's the interconnects between countries that are probably the biggest problem.

    --
    I used to have a better sig but it broke.
    1. Re:It always will be fragile by brunes69 · · Score: 3, Informative
      Take, for instance, the connections running between Europe and America. I bet most of them run in almost exactly the same place on the sea bed because it's the cheapest / shortest path to take. A fairly localized geological disaster (at least in geological terms) could cut all the cables at once; or at least enough to make to difference.

      This isn't a good example, because in this case most traffic would automatically be re-routed to go through Asia and the trans-Pacific cables. And if those went down it would go over South America Oceana.

      It would get much slower, sure, but would not cause an outage.

      There is no *technical* reason this peering relationship breaking down should be causing an outage either. If the both also peered with some third party that could service them both, like MCI or something, then the traffic would still get through. The companies are just being bull-headed.

    2. Re:It always will be fragile by jmilne · · Score: 1

      Engineers do actually consider redundancy when designing these things. Take, for example, TAT-14. They have a northern route that goes up to Denmark, and a southern route that lands in England. And that's just one of many fiber optic connections across the Atlantic.

      Distance is important, but so is redundancy.

    3. Re:It always will be fragile by squoozer · · Score: 1

      I have no hard data to back this up but IIRC when 9/11 happened it took out a major Internet switch center as well. Everyone talked about how traffic would get routed from Europe to America via Asia and the Pacific links but the reality was that most of America was unreachable from Europe for quite a few days. I suspect the places that were reachable had European mirrors. While there was a viable route it wasn't able to cope with the increase in traffic so basically every packet timed out.

      --
      I used to have a better sig but it broke.
    4. Re:It always will be fragile by MegaFur · · Score: 1

      "America was unreachable from Europe". Was Europe unreachable from America? Because I remember successfully connecting to http://news.bbc.co.uk/ to get some news about 12-14 hours after the twin towers were hit. (I live in MO, USA.)

      Also, no later than the very next morning, I was watching streaming video from the BBC about the disaster--and I was on lowly dial-up access as well. Possibly whatever location I was downloading from was a semi-local mirror, but the (live, current) data had to be getting from its source in the UK to the semi-local download point somehow.

      I think the people in this thread are weirdly defining "robust" to mean "absolutely no network quality degredation at all" and are defing "fragile" to mean "any kind of slow down whatsoever". I do not think those are good definitions for those terms. You wanna talk fragile? Let's talk thinnet, now that crap's fragile--any host in the chain gets disconnected and you lose the whole LAN. And then we have {br,t}oken ring--now *that's* fragile. :-)

      --
      Furry cows moo and decompress.
    5. Re:It always will be fragile by onepoint · · Score: 1

      I happen to disagree with you. I lived in the area affected and had no outages what so ever. outbound to europe, and I was in contact to my european counterparts durring this event and never encountered an issue

      When the area was destroyed, the major hubs of Newark ( one of the top 3 most fiber and wired location in the USA at the time ) and Westchester ( white plains i think ) took the entire load.

      the only people that lost connectivity were those ( isp's)that were in the peering location in WTC ( a few local isp's )

      if you lost connectivity with Europe, then it was really that they had there servers in WTC.

      Onepoint

      --
      if you see me, smile and say hello.
    6. Re:It always will be fragile by gskouby · · Score: 1

      Your statement of "If they both also peered with some third party......" would turn that third party into a transit carrier and not a peer. Transit = expensive; peering = cheap (relatively)

    7. Re:It always will be fragile by WuphonsReach · · Score: 2, Interesting

      Ah yes, the joys of thinnet. OTOH, it was very easy to debug if you knew how the thinnet was routed from cubicle to cubicle. When you had a broken segment, you went halfway down the line and terminated it off. If the segment started working, your problem was farther away from the bridge (repeater?). Otherwise, you would head back upstream towards the head of the segment and try again. Where you typically ran into trouble were users who constantly moved equipment (test labs, laptop users). User training fixed most of those issues due to the informal posse of coworkers who would hunt down the frequent offender.

      The previous topology in that office had been thicknet (where you had to manually tap the cable). Thinnet was seen as better. Or at least easier to build a network out of in a cubicle environment.

      Token Ring wasn't all that bad. Unlike thinnet, the physical wiring was more of a topology like today's ethernet where you had a dedicated cable running from the patch panel to the workstation's network jack. At least, it was wired that way in the buildings where I've seen it. So it was easy enough to plug/unplug stations from the network in a central location. The topology was also designed to deal with a single break (the stations before/after the break would loopback).

      The usual problems we had with TR were the fragile connectors (problematic for test environments / laptop users with frequent plug/unplug). Plus the issue that you only had 4Mbps (later 16Mbps) and a 4Mbps card wouldn't work on a 16Mbps network. Ethernet hubs/switches did a much better job of handling the upgrade path automatically where one port might be 10Mbps another 100Mbps and a third port running at 1Gpbs without redoing your entire network topology.

      --
      Wolde you bothe eate your cake, and have your cake?
    8. Re:It always will be fragile by 2008 · · Score: 1

      The beeb has a New York data centre, with a variety of US mirrors.

      Network map.

      --
      I quit!
    9. Re:It always will be fragile by pyrrhonist · · Score: 1
      but the reality was that most of America was unreachable from Europe for quite a few days.

      This isn't true. At the time, I was working for a company in the northeast that was working with Telecom Italia. We had no issues exchanging e-mail with them on 9/11.

      --
      Show me on the doll where his noodly appendage touched you.
    10. Re:It always will be fragile by beebware · · Score: 2, Informative

      Oh, for the record, the BBC has a brilliant network in the UK and the US - http://support.bbc.co.uk/support/network/ . I believe, although I haven't even attempted to confirm it, that they have peering agreements with most of the major UK ISPs.

    11. Re:It always will be fragile by MegaFur · · Score: 1

      That's neat. But how was I able to visit their UK site and get current info, how were they updating the New York data center if the line to Europe was down?

      --
      Furry cows moo and decompress.
    12. Re:It always will be fragile by 2008 · · Score: 1

      Well, firstly, the news was actually happening in the US so I imagine the US correspondents were updating the US BBC site directly. The link with Europe was probably either something simple like telephone, or via one of the satellites they have access to and use for remote cameras etc.

      --
      I quit!
  23. Keep it private by dada21 · · Score: 2, Insightful

    If you make the Top Tiers a government-controlled service, expect long term problems like censorship, taxation and regulations on sub-level tiers.

    Neither company involved in this dispute wants to do t is. They need to work it out, or other companies will find a solution and take the customers.

    If you're desperate to provide data to multiple top tiers, pay for a host that is connected to multiple backbones.

    There is zero need to mandate anything. Let the free market provide and we'll be safer in the long run. Let government provide and we'll see a slowly creeping tyranny online.

    1. Re:Keep it private by nagora · · Score: 1
      If you make the Top Tiers a government-controlled service, expect long term problems like censorship, taxation and regulations on sub-level tiers.

      All of which can, and have been, imposed anyway.

      Let the free market provide and we'll be safer in the long run.

      There is no such thing. It is an impossibility. Communism and Capitalism both have these utopian ideas at their hearts; that's why neither of them work in practice the way they're supposed to on paper.

      TWW

      --
      "Encyclopedia" is to "Wikipedia" what "Library" is to "Some people at a bus stop"
  24. Not a redundancy issue... by boldtbanan · · Score: 4, Interesting

    As I understood the problem, redundancy wasn't an issue. Level 3 was actively filtering out request to Cogent, however they came in. The redundancy was working, but Level 3 was playing NetNanny and blacklisting all Cogent IPs.

    1. Re:Not a redundancy issue... by Comen · · Score: 1

      Saying Not a redundancy issue must be true, all these people yelling about teh Net's Fragility sounds like BS to me.
      This must be one of the 2 providers are being assholes.

      When the peering of these 2 goes down, BGP should see these routes coming in from any other peers that these 2 are connected to, things might not be a fast between them as it used to but it should go down.
      Personaly if L3 did block these peoples ips I think that is total BS and would almost think you could sue or something.

      Even if peering is gone Cogent deserves the same treatment in terms of routes on the L3 routers as any other company.
        If L3 did this to everyone they wouldnt have any internet access.

    2. Re:Not a redundancy issue... by anticypher · · Score: 1

      Yes, blackholing is exactly what's happening. I have clients on both carriers in this dispute, and its really clear there is blackholing going on from both sides. Level 3 is being really obvious about it, because I doubt they have an experienced network engineer left working for them. Cogent has done this at least three times in recent memory, with AOL, with LambdaNet just before the assfu^Wbuyout, and with OpenSew^H^H^HTransit.

      If all the world's BGP sessions carried the whole routing table, and carriers were forbidden from creating blackholes, nobody would have noticed this childish display of stupidity. Routers would have routed around the damage. But with TWO major carriers blackholing routes, it is difficult to fix the problem. This is not a call for legislation (which I fear, now), because full BGP tables are rarely needed, and blackholing has many engineering functions.

      There are many small companies in Europe with an AS number, one cheap transit provider, and a few connections on local peering points for regional traffic because usually more than 50% of their traffic is local. When I say cheap, I of course mean Level 3 and Cogent, the two market bottom feeders.

      It has been a lucrative week for me, because I have the knowledge to fix this, and have been doing so to the point of exhaustion tonight (its 4:00AM, I've been up since things broke thursday, and my typing this reply was interrupted by a call from another potential client). I'll be buying the Level 3 and Cogent guys drinks when I get a chance, they've really fucked up their markets with this stunt, but I'm cleaning up.

      the AC

      Ooops, I think I displayed a little bit of editorial opinion in this ra^H^Hpost. how the hell did it get to 4 long paragraphs? Bedtime!

      --
      Hemos is like...sci-fi fans;he thinks technology is cool, but he hasn't bothered to understand the science it's based on
  25. This is so strange... by rubberbando · · Score: 0, Redundant

    I didn't even know anything about this until Howard Stern started talking about it this morning on his show. He was pretty pissed about not having email for 2 days. He also said that nobody informed him that it wasn't on his end. He was going nuts trying to get his computer fixed, calling every computer guy that he knew to come look at it.

    --
    DEAD DEAD DEAD DELETE ME
    1. Re:This is so strange... by cloudmaster · · Score: 2, Funny

      Knowing that it pissed Howard Stern off and wasted some of his time, I now feel much better about this outage.

  26. Governmental Role? by slipnslidemaster · · Score: 1, Interesting

    I don't believe in large government but this strikes me as one of those things that government is good for.

    Why wouldn't it be a good thing for some governmental agency to regulate the physical location of various installations? It seems to me that many providers use the same colocations to house their equipment. It would seem smart that there would be some regulation to prevent all the Internet eggs in one basket.

    Create several more physical IXP's that are located in geographically diverse areas with redundant connections. Then regulate that only a limited number of companies could colocate together within a certain number of IXP's.

    This could prevent one companies "disagreement" with another from effecting the traffic being routed to an alternative link.

    Does this make sense?

    --


    "What the hell is an aluminum falcon?"
    1. Re:Governmental Role? by pturing · · Score: 1


      Does this make sense?
      Oh, I'm sorry, now.
      But we do have a parting gift for you.
      [insert gote-see link here]

    2. Re:Governmental Role? by Anonymous Coward · · Score: 0

      It's goatsex, moron.

      the domain name was goatse.cx read phoneticaly that is goatsex.

    3. Re:Governmental Role? by pturing · · Score: 1


      gorram it, I will mispronounce things how I want

      [insert goat-seeks link here]

  27. bidding-based protocols by G4from128k · · Score: 1
    Perhaps routers (both the devices and the companies) should bid for packets in a real-time (or with a periodic) reverse auction. Rather than count segments to the destination, the logic would minimize the cost-to-deliver the packet. For most connections between true peers, the total of the charges would be zero.

    Bandwidth isn't (and never will be) truly free as long as the hardware and admin labor has a cost. But if we seek way to deliver the most packets at the least cost, then market forces will drive the price down. The total system would route around both damage and inefficient (= high cost) parts of the network.

    --
    Two wrongs don't make a right, but three lefts do.
    1. Re:bidding-based protocols by ultranova · · Score: 1

      Perhaps routers (both the devices and the companies) should bid for packets in a real-time (or with a periodic) reverse auction. Rather than count segments to the destination, the logic would minimize the cost-to-deliver the packet. For most connections between true peers, the total of the charges would be zero.

      This is a good idea. However, it demands an untamperable log system - that is, you need to be able to prove that router X sent you n packets, and the owner of router X needs to be able to prove that he sent no more than n packets. Also, everyone needs to be able to own all the hardware they operate - that is, owner of router X has full control over his equipment and none over yours, and you have full control over your equipment and none over router X (no DRM of any kind, in other words).

      One possible way of achieving this would be to use public key cryptography. Owner of router X generates a public-private keypair, and enters the public key to the contract (and you enter the public key of your router, as well). Then, whenever X sends you a packet, it calculates a hash for that packet, and then signs the hash. When the packet reaches you, you check that the hash and signature are correct. If they aren't, just discard the packet; if they are, forward the packet and store the hash-sig pair. Then, when billing, each hash-sig pair means a received packet; you can easily prove that they were indeed signed by router X, since the public key of X is in the contract; and you can't forge them, since you don't know the private key of X.

      Now, there will be problems if two packets are exactly identical, since they will have exactly identical hashes then too; that can be easily solved simply by adding a running serial number into the packet wrapper frame (which you need anyway to send the hash and sig). That only leaves hash collisions to worry about.

      Another way would be to forget hashes completely and just sign a running serial number token, with one sent along with each packet. That would get rid of collisions completely, and be much faster since you wouldn't need to hash the packet. However, I'm not sure if signing billions of sequential integers with the same private key would be cryptographically sensible - does it make the private key any easier to deduce ? Any cryptography experts here ?-)

      Since the length of a packet can vary, the serial-sig format should propably be made into a serial-size-sig triplet. This would allow the calculation of actual bandwith usage, and not just the amount of packets transmitted (but of course you can still get the latter too).

      Finally, what happens when a router is overloaded and starts dropping packets ? Or suppose a router accepts a packet, logs the token, and then the target router goes down in the middle of a packet send. The packet delivery has been paid for, so can the packet just be dropped ?

      Bandwidth isn't (and never will be) truly free as long as the hardware and admin labor has a cost. But if we seek way to deliver the most packets at the least cost, then market forces will drive the price down. The total system would route around both damage and inefficient (= high cost) parts of the network.

      That, or the ISP's will put even stricter limits on the amount of traffick you can generate before being cut off.

      --

      Forget magic. Any technology distinguishable from divine power is insufficiently advanced.

    2. Re:bidding-based protocols by InfraRED · · Score: 1

      there is no need for fancy untamperable logging at all.
      If some peer does not honor payment obligations, its trust drops and won't be able to communicate.

      --
      metamoderate!
  28. Re:Ask Slashdot by AlexTheBeast · · Score: 3, Interesting

    The problem with web services is that they need for the internet to be completely secure and completely reliable. The internet of today is neither.

    Physicians trying to use the internet to take care of critically ill patients are already experiencing this. Radiologists sitting home reading films are seeing this as well.

    Is 100% on neccessary? Hell, VoIP is making money like crazy over this unstable network of ours.

    My suggestion is to test with people that will understand the limitations of your service. Then get a little VC money to spread your servers out.

  29. The fragility of the net by elfguygmail.com · · Score: 5, Informative

    It's very true, and anyone can see how a few big companies basically make the net work in north america. Simply do traceroutes to various big web sites, and you'll notice the packets always go across the same networks. The biggest one seems to be alter.net (MCI), with others including Level3, above.net, AT&T and UUnet. Basically you remove any of these and the North American part of the Internet would be in chaos. The problem is because most ISPs do the same thing. They pick a primary provider, and get a backup one. The problem is they all pick the same few primary companies, and their backup links are much smaller pipes.

    1. Re:The fragility of the net by lostlogic · · Score: 1

      That's not entirely true, for that matter, traffic from my home server to my leased server takes a route via sprint, and traffic in the other direction takes a route via telia (both have one end in ALTER.NET). Any good hosting provider, and any really good internet service provider should be truly multihomed.

      The point being that as others have pointed out, you simply get what you pay for. If you buy cheap service, you get a cheap single homed connection.

      --
      --Brandon
    2. Re:The fragility of the net by laffer1 · · Score: 1

      Yes, all isps should be multihomed. I worked for an isp for a few years as a sys admin. We only had 4000 customers (hosting and dial up), but had two upstream providers. At the time we had a T3 to UUNET (MCI) and an ethernet link at the telco to verio which had a T3 to detroit from that location. Our modem bank was at the telco and so dialups routed through verio most of the time. We had 3 t1 circuits to our office which was about 10 miles away. The network topology wasn't the best, but it did work. We had several occasions when one of the links went down and we just adjusted some routes to bring everything up. If a small isp can do this, then i think all should.

  30. TWC by LividBlivet · · Score: 0, Redundant

    Time Warner Cable in upstate NY was unable to connect to several sites. SETI and Newsfeeds notably on Wed.

  31. Based on expected traffic routed thru each network by Anonymous Coward · · Score: 1, Informative

    If the traffic is about even both ways, the peering aggrements are made with no cash exchange. If it's uneven, the network not bearing its fair share of traffic usually has to pony up some cash as part of the "peer" aggreement. If things don't turn out as expected, the network carrying an unfair burden will usually back out or renegotiate the peer aggreement. You usually don't find out what the actual network traffic is until you start peering.

  32. ah peering by bigpat · · Score: 3, Interesting

    The only time peering should involve an ongoing exchange of money for bandwidth should be when a network is primarily serving as an intermediary between other networks, such as long haul or backbone networks.

    But if most of the traffic from other networks is going to customers that are connected and already paying for your network's service then it makes no sense and is simply wrong for a network to start charging other network providers. It breaks the end to end communication model and is providing your customers with less than the service they are paying for. People pay for internet connectivity so they can transfer data between other users on the internet, not just the ones on your company's network.

    If money exchanging hands is at all appropriate in this case it might be for the actual installation of routing equipment which establishes the physical connection between networks.

    1. Re:ah peering by Anonymous Coward · · Score: 0
      I couldn't have said it better myself.

      If someone on say... a RoadRunner cable modem has already *paid* for Internet service and RoadRunner has in turn already *paid* L3 for their fiber for their service on behalf of their customer base, and likewise a web site has already *paid* Cogent for hosting their server, why should Cogent in turn pay Level3 again when L3 was already paid?!

      If this nonsense continues, it would be perfectly reasonable we may encounter warning pages in certain segments of the Internet like this:
      "WARNING!
      YOU ARE REQUESTING CONTENT IN A PORTION
      OF THE INTERNET FOR WHICH YOUR ISP
      HAS NOT PAID TO CONNECT.

      PLEASE ENTER YOUR CREDIT CARD NOW TO ACCESS."
      (I wonder how this could work with email since there's no chance for a warning page. Guess emails just bounce....;-)

      Kidding aside: For those of you who don't quite gather the scope of this: Georgetown University and (as mentioned by the article cited above) The Museum of Fine Art in Boston are just two example of organizations *heavily* affected by this. Oh, and there's no resolution on the horizon, either.

      For those of you that haven't already done so, you might consider a trip to your local Attorneys General web site to file a formal complaint.
    2. Re:ah peering by ranson · · Score: 1

      Actually the Tier 1 providers (read: backbones) have settlement-free peering agreements anyway so no bits are talled to be billed, and no money exchanges hands for peering with each other.

      > The only time peering should involve an ongoing...
      > exchange of money for bandwidth should be when a...
      > network is primarily serving as an intermediary between...
      > other networks, such as long haul or backbone networks.

    3. Re:ah peering by Anonymous Coward · · Score: 0

      If this is true, then I'm going to call my house Home Networks Inc., and I want a free peering relationship with AT&T, MCI, Level3, Cogent, and any other player. Since the Internet is all about the people on the Internet wanting to talk to me and my family, then these companies should *want* to peer with me? Right?

      The point is that if one company is much larger than the other, then the smaller company gains more benefit from the peering arrangment. Simply put, the smaller company gets better access to 10 million endpoints while the larger company only gets better access to 1 million endpoints.

      User-base is considered an asset. If I have lots of users that you want to talk to, there's a good chance that you're going to be willing to pay to talk to them. Especially if one of my users is /.

      Just my $0.02

      -AC

    4. Re:ah peering by bigpat · · Score: 1

      Lets take your example of a home.

      Do you really think that $30 a month is really paying for the "asset" of connecting to the rest of the world with its millions of users. By your example, the person paying the most would be the end user because they have the most unequal relationship and bring the least to the table. But I don't pay 100 million dollars per month to connect to Comcast.

      I think it is far closer to my example, We merely pay for the cost of our connection to the network and the cost of maintaining long haul backbone networks which connect to other networks.

  33. Re:Net blackouts by crazygeek02 · · Score: 1

    No it can't. Think before you post.

  34. not a blackout by bradk500 · · Score: 2, Interesting

    All this crap about it showing weakness in the internet is uninformed bs. They didn't just stop peering, but they are actively blocking traffic from cogent. If Level3 had just stopped peering the traffic would reroute around the problem. The only time you will see problems is if your a cogent customer trying to get a single homed computer on level3's network. We are a cogent customer and an internap customer, and to get around the problem I just reouted traffic destined for level 3 networks over one of our internap t's. This solved the issue for us.

    1. Re:not a blackout by Loceur · · Score: 1

      Cogent and Internap? That's an aweful mix.

      That's like mixing a fine port and Lonestar beer.

    2. Re:not a blackout by bradk500 · · Score: 1

      Yes your 100% correct, we are in the process of cutting over. This just kinda speeded it up. Also goes to show that internap is worth the extra money for just such issues like this. We haven't even finished cutting over and the extra cost has already been justieifed many times over as would have lost a weeks worth of testing with one of customers who use level3.

  35. So the internet is breaking down by iminplaya · · Score: 2, Insightful

    Privatization strikes again. You put the infrastructure into the hands of a few powerful people and this is what you will get. Those big power outages happened for the same reason. We aren't holding those in charge responsible. There is no redundancy when there is only one provider. They can cut you off and what are you going to do? Only community services and coops can provide the necessary robustness. But it seems to be more convenient to just hand it over to corporate pirates.

    --
    What?
    1. Re:So the internet is breaking down by the_real_bto · · Score: 1

      "Privatization strikes again. You put the infrastructure into the hands of a few powerful people and this is what you will get." Are you arguing that government control moves power from the few to the many? That is backwards to my way of thinking. The quickest way I can think of to concentrate power is to put the government in charge of it.

    2. Re:So the internet is breaking down by the_real_bto · · Score: 3, Insightful

      "Privatization strikes again. You put the infrastructure into the hands of a few powerful people and this is what you will get."

      Are you arguing that government control moves power from the few to the many? That is backwards to my way of thinking. The quickest way I can think of to concentrate power is to put the government in charge of it.

    3. Re:So the internet is breaking down by iminplaya · · Score: 1

      Well, seeing that we are responsible for the actions of our govt, and it is suppose to represent us(if it doesn't, you know what to do), I can't think of a better place to concentrate that power. In reality, it disperses it amongst all of us. If the govt is out of control, it's because we let it happen through our own negligence. It's because we aren't vigilant. The solution is obvious. The only obligation a corporation has is to maximize profits. And we all know they will do whatever it takes to accomplish that.

      --
      What?
  36. The Internet doesn't match its sales brochure. by nuggz · · Score: 1

    The complicated, interlocking nature of networks
    But when you choose to have a single critical link you don't have an interlocking web of connections.

    even if a critical link is broken.
    If it was a web there would be no critical links to break. The problem is that for various (technical/economic) reasons there is a backbone (or series of backbones) it isn't really an interlocking web.

  37. The peering problem is bad.... by doormat · · Score: 1

    when you hear Howard Stern complain about not getting emails and whatnot.

    --
    The Doormat

    If you're not outraged, then you're not paying attention.
    1. Re:The peering problem is bad.... by Anonymous Coward · · Score: 0

      If Howard Stern complaining means a problem is bad, then holy hell, the world must be coming to an end.

  38. back in the day.... by SQLz · · Score: 1

    Back in the day in there was no 'peering' contract. It was all techies helping techies and peering arrangements would happen with a handshake over some good sushi. You take my routes, I'll take yours. Then all the lawyers got involved, its been a nightmare ever since.

  39. WiFi or WiMax by leon.gandalf · · Score: 0

    Here is a solution, set up your wireless router to bridge with a friend or someone you know. Keep this up untill you have a realy big WAN. Hell after a while you could have "INTERNET 2.5 - Wow that latency is a BITCH. But the man can't keep me down!"

  40. re the fragility of the internet by argStyopa · · Score: 1

    Hey, but you Euro/UN types go ahead an slap down your new root servers wherever you want....that certainly won't screw everything up.

    http://politics.slashdot.org/article.pl?sid=05/10/ 06/1241227&tid=95&tid=219

    --
    -Styopa
  41. Hmmm.... by Hrvat · · Score: 1

    If I remember correctly, shouldn't network protocols be flexible enough to find another path to the server? Let's say that you are on one network and the desired server is on another network that have their direct connection killed. Can't you go and find an alternate route through a peer that is connected to the both of you? It would be slower but shouldn't this happen?

    --
    TANSTAAFL
    1. Re:Hmmm.... by Anonymous Coward · · Score: 0

      -5 for not paying attention.

      The problem is not that the routing protcols have failed. The problem is that rather then simply cutting the peering link, they've adjusted their routers to DROP traffic on the floor that was supposed to go over the old peering link.

      Basically, they're saying... "sure, we'll send your data to XYZ" and then turning around and dumping it in the dumpster out back.

      (I guess you could look at it as an example where the routing protocols can get separated from reality. IOW, the routing protocols have no way to determine whether a particular router has lied to them about being able to handle a particular route.)

    2. Re:Hmmm.... by Sedennial · · Score: 1

      Yes, but if you read the article you would find that this only affect customers who are single homed behing Cogent OR L3 for that very reason. If you are singlehomed and your ISP shuts down all their peering and no longer accounts route announcements containing L3 AS (or vice-versa) you are fundamentally screwed.

      This is a classic example of why all businesses who are internet dependent (most busineses nowadays) should either have high-speed DSL as a backup at the least, or some other provider ready to go. A number of ISP's will provide 'standby' connectivity for a nominal fee where everything is ready to go and you only pay a maintainence fee unless you need to actually turn up the connection.

      You get a cheap(er) router for your small business, get private BGP (announce private AS) between two isps and yourself, and only use the second one if you have a situation like this.

      We have a number of customers we work with that we have helped them set up exactly this scenario. Business cable, DSL, and fiber customers who have a 2nd ISP (usually 3-5 Mb DSL) they only use for fail over. Usually costs under $75/mon to have the backup service ready to go.

    3. Re:Hmmm.... by Hrvat · · Score: 1

      Ok, so they didn't only shut down the link, they actively will not route anything that resolves in the other network?

      In any case some business have no alternatives especially if they are in a rural area with one or two ISP's on the same backbone.

      --
      TANSTAAFL
  42. Re:Net blackouts by Anonymous Coward · · Score: 0

    To the moron mod, that was FUNNY! I'm pretty sure the author meant it as a joke, because I laughed when I read it. Flamebait? Sheesh.

  43. Blackhole by TallMatthew · · Score: 1

    If you were ISP in distress, wouldn't you find a way to circumvent this downed peering point? I'm sure other Tier 1s have jumped in and offered transit for a modest sum. If I were an evil ISP and I wanted to make sure that didn't happen, I'd blackhole all the netblocks (or any ASPATH with his AS) from the guy I wanted moolah from until he forked over the funds. Unless I'm missing something, that's the explanation for why networks who are purchasing transit from Cogent/L3 are having trouble.

  44. Cogent Sucks by Lamont · · Score: 2, Interesting

    As a customer who has had Cogent inflicted on us (when Verio sold all their domestic internet lines to Cogent), we've had nothing but pain and bumbling inefficiency from them for the last six months.

    I contacted Cogent's "premium" help desk last night when I found that I was suddenly no longer able to get to our networks in Australia. The tech had no idea that his own company was in the middle of a huge peering battle with L3. I had to tell them!

    1. Re:Cogent Sucks by /dev/trash · · Score: 1

      Oh no. I bet you felt superior telling the minimum wage single Mom that she's ignorant.

  45. Mod parent insightful by Anonymous Coward · · Score: 0

    nuttin here

  46. North Pole Run coming soon by davidwr · · Score: 2, Funny

    As soon as all that pesky arctic ice melts away, it'll be cheap enough to run cable across the pole.

    As a bonus, Santa's new underwater toy factory can tap into it.

    Woo-hoo, faster email to Santa! Hope the jolly old elf doesn't discover online pr0n or he'll never get those presents made on time.

    --
    Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
  47. Re:Net blackouts by Cholobos · · Score: 0

    Maybe a :) would have made it easier for you to understand.

  48. Quite Fitting! by Anonymous Coward · · Score: 0

    This "homemade" MP3 is quite fitting in this situation. A must have for all concerned. :-) http://www.wps.com.au/music/Rumors.mp3

  49. This was predictable by PhilipPeake · · Score: 4, Interesting
    The Internet was designed to be resiliant to malfunctions and automatically take appropriate action to ensure connectivity.

    Unfortunately, that is not the Internet that we have today. In the original Internet, every router knew about every network connected to the Internet. Most networks had connectivity to many other networks. Discovery protocols allowed alternative routes to be discovered if one failed.

    Today, we don't have a (mostly) fully connected net, we have ISPs who don't know anything about networks which they don't "own", only that certain IP prefixes need to be passed to ISP x, y or z.

    This makes the infrastructure much more fragile than it was originally intended to be. We ended up with this for a few reasons. First, the wimpy routers in use at the time had limited memory available to hold the network maps. The answer chosen was to no longer attempt to hold a full world view, but to divide the world into regions, certain IP prefixes would "belong" to those regions, and all any router would need to know about was networks in its region, plus how to route traffic to other regions, who would take care of routing within the region. This led to "backbone" connections - high capacity links needed because all traffic between regions now didn't "diffuse" through the network, but was channeled into specific connections. It also set the scene to allow the net to be commercialised, those regional centers were obvious "choke points" that an enterprising company could own and pretty much dictate the pricing to lower level enterprises who would do the dirty work of dealing with end-users.

    Slowly but sureley the Internet evolved into a system dependent upon a few companies with high-speed links between them - prime candidates BTW, as locations for government control to be imposed. The self-healing nature of the original Internet was lost because all traffic HAS to pass via the top level companies infrastructure and over their interconnect backbone connections.

    The "self healing" Internet is long gone.

    1. Re:This was predictable by Eli+Gottlieb · · Score: 1

      There's also the fact that if every seperate computer were assigned a single IP address (rather than a public IP for its network and a private for the machine itself) we'd all have to switch to (GASP!) IPv6, and for some reason nobody wants to do so.

    2. Re:This was predictable by Red+Flayer · · Score: 2, Interesting

      Slowly but sureley the Internet evolved into a system dependent upon a few companies with high-speed links between them - prime candidates BTW, as locations for government control to be imposed. The self-healing nature of the original Internet was lost because all traffic HAS to pass via the top level companies infrastructure and over their interconnect backbone connections.

      This is what happens when you have an industry based upon a high cost of entry (physical infrastructure, here) and a low marginal cost of supply. We need fat pipelines because we demand fast speeds and high volumes for our traffic. If we didn't have regions, but instead had the "original self-healing internet," how long do you think it would take to download big files if the source didn't happen to be just 2 or 3 routers away? Say goodbye to streaming video, etc.

      Net cost of transmission would be far higher for packets that are many routers away in a truly web-based system, since not all apths are equal.

      The problem is, how do we balance cheap efficiency (fatline "superhighways") with expensive redundancy to optimize the system for all participants?

      --
      "Trolls they were, but filled with the evil will of their master: a fell race..." -- J.R.R. Tolkien on Olog-hai
  50. Re:The small should pay for the big? (mod this up) by gskouby · · Score: 5, Informative

    About 4 months ago I got a call from a sales critter at Cogent saying "We will knock 50% off of the price you are paying for your L3 connectivity if you drop them and come be our customer." I was kind of surprised at the boldness of this proposition because they were specifically targeting current L3 customers. I was even more surprised to find out from others that this sales pitch from Cogent was company wide. Of course this pissed off L3 and that was the start of this pissing contest.

  51. Re:The Truth is by Anonymous Coward · · Score: 0

    Cogent is dumping a lot of data on Level 3 in a peering arangment. Which means that the data is destined for customers on Level3's network. Since Level3 makes money from charging customers for bandwidth, this is a benefit. More data to/from their customers means more money. Level3 however, wants to make money on the bandwidth twice for each customer (once by charging cogent to talk to Level3 customers and again from Level3 customers).

    The only way this would be a problem is if cogent is using the connection as a "transit" connection, routing data for other networks through Level3 which doesn't seem to be the case.

  52. Crazy Idea by Luyseyal · · Score: 1

    OK, I have a crazy idea. I am NOT a net admin and am largely blue skying here.

    Damage: Level3 won't accept Cogent traffic.

    Horrible hack: tunnel BGP traffic to Level3 customer who masquerades requests as local traffic.

    Yeah, the real solution is tier 2 folks having more peerings, but as a nasty workaround is that hack feasible?

    Can you tunnel BGP traffic in TCP or ssh or something?

    -l

    --
    Help cure AIDS, cancer, and more. Donate your unused computer time to worldcommunitygrid.org. Join Team Slashdot!
    1. Re:Crazy Idea by gfilion · · Score: 2, Insightful

      Damage: Level3 won't accept Cogent traffic. Horrible hack: tunnel BGP traffic to Level3 customer who masquerades requests as local traffic.

      You don't need to masquerade anything, if you're connected to Level3 and Cogent, just configure your router to advertise your route to the Level3 network on the Cogent side and vice-versa.

      Then watch your router melt under the hundreds of gigabits of traffic -- that you'll have to pay for both ways. Congratulation, you're the new peering agreement between Level3 and Cogent!

    2. Re:Crazy Idea by mihalis · · Score: 1

      You don't need to masquerade anything, if you're connected to Level3 and Cogent, just configure your router to advertise your route to the Level3 network on the Cogent side and vice-versa.

      Then watch your router melt under the hundreds of gigabits of traffic -- that you'll have to pay for both ways. Congratulation, you're the new peering agreement between Level3 and Cogent!

      Very well explained, thank you sir!

      I wonder if all the other Tier 1 providers together could simultanously announce routes connecting L3Cogent with equal metrics and survive. Then bill both Level 3 and Cogent for net transit on all their traffic. It seems to me that by their actions both L3 and Cogent are not currently Tier 1 and deserve not to be treated as such.

    3. Re:Crazy Idea by mihalis · · Score: 1

      Sorry that got mangled, I wrote "simultanously announce routes connecting L3" less than sign, dash, greater than sign "Cogent". Also it's spelt "simultaneously". Sorry!

  53. it does not matter by Anonymous Coward · · Score: 0

    nobody noticed because nobody uses cogent or level three.

  54. Re:Net blackouts by Evil+W1zard · · Score: 2, Insightful

    Ah I see so by giving control over to the UN it will magically put in place all the hardware, software and correctly configure the web to never fail? Hopefully this statement was made to be joke because otherwise it doesn't make a bit of sense. I picture the web now as a 100,000 foot long giant slinky that someone has twisted into oblivion. I don't even know if the web can be fixed at this point...

    --
    News Reporters Make Tasty Polar Bear Treats!
  55. You have to define 'size' by oneiros27 · · Score: 1

    Typically, when you're dealing with peering, it's the amount of traffic that you're pushing on the other guy, because people on your network want to connect to places on their network. (or use your network to get to another network that you peer with)

    So, when a mom & pop ISP connects to one of the big guys, there's very little of interest on their network as an endpoint, and they probably don't advertise better routes than the tier1 already has, so they have to pay for the priviledge to be connected.

    Now, in the case of Cogent, if I remember my ISPs correctly (I've not worked for an ISP for 6 years now), they were PSI, which merged with a few other networks -- the problem was that the networks weren't well connected, and so they were accused of what was known as 'hot potato routing' -- trying to get the traffic off their network as quickly as possible, even if it wasn't the shortest number of hops to the destination.

    Some ISPs will also pad the routes that they advertise, so that they basically tell other people 'we'll take our own traffic, but we don't want to route other people's traffic through us'. (or it may be that their internal networks are so badly interconnected, that it really a whole lot of hops -- I mean, what good does it do to advertize that you connect to the moon, if you're going to route everyone by way of Jupiter?)

    I'm guessing that Level 1 was montioring the traffic flowing between themselves and Cogent, and noticed that something seemed significantly amiss, suggesting that Cogent was taking advantage of their services.

    I don't know the specifics of this case, however. so this is all conjecture, based on 6+ year old knowledge of ISPs. (and well, I can't find what happened to Boardwatch, so I can't pull up nice maps of the big players)

    --
    Build it, and they will come^Hplain.
  56. That's why the UN should run it! by donstenk72 · · Score: 1

    This is it. We're coming to take over the net from you guy's. Your companies are obviously not able to run it reliably. All non US ISP's should be required to resolve DNS to un.un.

  57. Would this affect everyone? by Peldor · · Score: 1
    Or just the customers of these two companies?

    Because my ping times were crap yesterday to the servers I normally play games on.

  58. Monitor it yourself by dereference · · Score: 4, Interesting
    I found this site while trying to research the problem. I wish I had known of it earlier; it provides a very nice (near) real-time snapshot of all the Tier 1 peering:

    http://www.internetpulse.net/

    I'm not affiliated with them in any way, and I'm sure there are other similar sites, but I thought it was worth mentioning.

    1. Re:Monitor it yourself by Cally · · Score: 1

      And what, pray tell, might you mean by "tier 1 peering"? Hint: the internettrafficreport matrix isn't showing you what you think it is, and tier 1 doesn't mean what you think it does. (It may in fact not mean anything, although informed opinion is divided on the matter...)

      --
      "None are more hopelessly enslaved than those who falsely believe they are free." -- Goethe
    2. Re:Monitor it yourself by Anonymous Coward · · Score: 0

      Could you kindly spell out what you mean by this, for those of us playing the home game?

  59. Pending Assingment by mattyohe · · Score: 1

    The Internet Health Report shows the interruption, and I noticed there is also the "Pending Assingment" as mentioned in these articles. I'm curious as to who this is.

    --
    - what is the definition of simultanagnosia?! I've been meaning to look it up!
    1. Re:Pending Assingment by gfilion · · Score: 1

      The Internet Health Report shows the interruption

      Your link doesn't work, use http://www.internethealthreport.com/

  60. I'm affected by AndrewR81 · · Score: 1

    I can't access my web host, textdrive.com, nor wikipedia.com from my resnet connection here at Waterloo University. Apparently they're relying exclusively on Cogent. It's quite frustrating because I can't can't do much. I have, however, found some anonymous browsing services and HTTP proxies that work well enough.

    I've been in contact with the Information Services Technology here and it looks like some more important services are being distrupted:

    Date: Thu, 6 Oct 2005 16:34:21 -0400
    From: Yuri Kolomiyets
    Newsgroups: uw.cscf.help
    Subject: Re: Can't connect to textdrive.com (or wikipedia)

    Oxford University Press is broken too, and library is actually paying to get
    access to those resources.
    Yuri Kolomiyets

    Network Support
    Information Systems & Technology
    University of Waterloo
    On one hand it's good for me that official services are affected, as it should boost priority.
  61. Tor Around The Block by _bug_ · · Score: 1

    I'm sitting off an ISP that plugs into Level-3 so this blackout is definately causing me problems. I can't get to Penny Arcade or Megatokyo (clearly this is a crisis). So I'm using Tor to get around it. Yeah, it's really slow at times, but it works.

    ( And as impossible as it sounds, there are more important sites I can't get to either, like the support site for a couple bits of software we use rather heavily at work. )

  62. so out of date by pturing · · Score: 1

    btw, this problem has already been resolved

      6 verio.dfw03.atlas.cogentco.com (154.54.10.30) 7.682 ms 7.723 ms 7.283 ms
      7 pop1-dls-p3-2.atdn.net (66.185.133.93) 7.515 ms 6.929 ms 6.793 ms
      8 bb1-dls-p0-0.atdn.net (66.185.133.80) 8.176 ms 66.393 ms 13.965 ms
      9 pop2-dls-p0-0.atdn.net (66.185.133.97) 7.083 ms 6.703 ms 7.292 ms
    10 rr-houston.atdn.net (66.185.132.18) 7.571 ms 7.125 ms 7.943 ms

    or at least it's been fixed for roadrunner cogent in Texas

  63. Cogent is Blocking its Own Sites by Anonymous Coward · · Score: 0

    The dispute between Cogent and Level 3 has been extended by Cogent itself. They are apparently blocking access to sites within their network based on the origin address. From a host on charter.com a traceroute ends at p12-0.core02.ord01.atlas.cogentco.com but from a host on ameritech.com the traceroute continues all the way to 66.28.205.104.

    The Cogent management appears to be deliberately creating a DoS situation for publicity purposes.

  64. Re:Net blackouts by Cholobos · · Score: 0

    Before I 'troll' more hateful comments, append a smiley to the first comment. It was a joke people, geez!!

  65. Sig: Key to financial independence... by dpilot · · Score: 1

    Watch your investments "fail", having run off with all your money.

    The real key to financial independence appears to be choosing your parents wisely.
    IMHO this is unfortunate, it can give rise to phenomena like "Paris Hilton."

    --
    The living have better things to do than to continue hating the dead.
    1. Re:Sig: Key to financial independence... by mjh · · Score: 1

      First, can I assume that you agree that spending less than you earn, and doing it for a long time are both very important aspects of financial independence?

      If so, then I absolutely agree with you that there's risk involved with investing. But there's also risk involved with not investing. If you spend less than you earn, what do you do with the difference? If you simply stick it in a safe and do nothing with it, the risk associated with this "strategy" is inflation. Your purchasing power will decline at roughly 3% per year. In other words, you lose when you do absolutely nothing. If you put it into a savings account that earns 0.5% per year, you still lose to inflation, but just not as much.

      It's a dramatic oversimplification to say that you can just "invest". Realistically that means building a well diversified portfolio, and maintaining it. But there's only so much I can put into /.'s 120 character sig limit.

      --
      Key to financial independence: Spend less than you earn. Save and invest the difference. Do it for a long time.
    2. Re:Sig: Key to financial independence... by dpilot · · Score: 1

      I used to save more than I earned, but that was in a previous phase of my life. I'm now known as the father of a college student, in a few more years to become the father of two.

      Actually, I'm not arguing against investing. I'm merely annoyed that there is a class of people that simplistically spouts your forumula as if, "That's all it takes to become rich." Some of that class of people are Ken Lay and whatsisname Fastou. Some others of that class ran the Silverado Savings and Loan.

      Right now I'm watching economic policies I disagree with produce a mediocre economy with cost of living increasing faster than my pay, my "investments" on about the same curve, and no change of course in sight. Perhaps one reason corporate America isn't doing as well as they like is that as they cut their costs on the backs of the American worker, they're destroying the ability of those workers to afford their products. Maybe the American worker is overpaid, but he has also been the strongest market in the world, at least until he's unemployed.

      I'm just grumpy. Ignore me.

      --
      The living have better things to do than to continue hating the dead.
    3. Re:Sig: Key to financial independence... by mjh · · Score: 1

      There are a lot of people who do bad things, but things that happen to be true. If Hitler had ever said e=mc^2, that would not have made the equation incorrect. Similarly, it doesn't matter that Ken Lay and Andrew Fastow encourage saving and investing. It's still good advice even if they're bad people.

      I would also encourage you not to give up hope too quickly. The american economy is very resilient. It's completely rebuilt itself countless times. In the process destroying old jobs and replacing them with new ones. This is another reason why saving is so important. If you have sufficient emergency savings, you can weather a short term loss of a job, and train yourself for a new one.

      --
      Key to financial independence: Spend less than you earn. Save and invest the difference. Do it for a long time.
  66. The problem here is conflicting business models by Mulligan · · Score: 2, Informative

    At the fringes there are really two types of internet service offered: upstream and downstream. Most consumers (individuals) need a lot of downstream and very little upstream. They typically are sold assymetric service that is heavily biased in this direction. My cable connection, for example, gives me ~5Mbps down and 768kbps up. On the flip side are the content providers who typically need a lot of upstream bandwidth and less upstream bandwidth. ISPs have found that these customer are willing/able to pay quite a bit more for their internet connections. Therefore, the law of supply and demand has increased the cost of connections with higher upstream capacity.

    Several levels up the ISP heirarchy, however, there are mostly only symmetric lines (T3, OCx, ...) providing equal upstream and downstream bandwith. In order to maximize the use of this bandwidth, many providers try to balance the number of content providers with content consumers in order to use the upstream and downstream capacity equally. In theory, this usage should be well balanced by the time it reaches the Teir 1 providers.

    The problem we are having right now is caused by Cogent not subscribing to that business model. They have found that the cost to support content consumers is much higher than the cost to support providers. (If for no other reason than there are far more of them.) So, their business model skews heavily towards the provider customers, reducing their operational costs. This, in turn, means that they are able to offer lower costs to those content providers -- in many cases undercutting the other big service providers such as Level 3

    This, of course, makes the other providers unhappy because it cuts into their high-yield business. So, occasionally, one of them demands compensation for "transit" instead of providing free peering. They do this because they feel (rightly IMO) that Cogent is able to make more money on these high paying content providers by using an asset owned by the other service providers -- the online customer/consumer base. Basically, Level 3 is telling Cogent that because Cogent is making money by using that virtual asset owned by Level 3, Cogent owes Level 3 some sort of compensation. It is worth noting that several other Teir 1 providers already take this approach with Cogent and Cogent is forced to pay for "transit" service to those providers' customers.

    As long as all the Teir 1 providers cooperate, the system works reasonably well. However, in this case, Cogent is trying to take advantage of that informal cooperation to make some extra money. So, they are being capatalists. In this case, capatalism is at odds with cooperation and the system is not working well.

    Many people are calling for government regulation to prevent this sort of situation. I expect this to cause some major problems. The issue could be resolved if all the Teir 1 providers would realize that there is a different market value for ingress and outgress traffic. In a free market, I expect that the ingress traffic (corresponding to upstream traffic of content providers from the lower levels) would have substantially more value than the outgress traffic (downstream traffic to consumers). The outgress traffic might even have negative value (meaning that a service provider would charge to take care of it). In the case that two peers balance their traffic well (the ideal cooperative solution) no money needs to change hands. In the other cases (like this one) the ISP with excess outgress usage should probably be charging the one with excess ingress.

    Unfortunately, there is no fluidity to the system between the true market (the upstream and downstream bandwidth consumers) and the core market (the Teir 1 providers). If there were, Level 3 could justify their demand for more money based on the value of the traffic they were accepting from lower down the food chain.

  67. have had cogentco problems since august by Anonymous Coward · · Score: 0

    I have servers on the backbone of a European Telco (which I don t name because they have shown displease at my complaints) which appears to receive traffic from the US via Cogentco. While normal web navigation isn t affected, since early august all ftp traffic directed to our servers going tru Cogentco (to my limited knowledge it is unfortunately ALL traffic coming from the US) is capped at speeds ranging from 10 to 24 kB, which therefore has been the maximum upload speed to our servers for over 2 months. This Euro Telco isn t Level3, and the trouble isn t limited to the past few days. Maybe there s more than the story of a temporary peering breakdown between Level3 and Cogentco. IS anybody experiencing bandwidth problems on US Europe traffic involving Cogentco ?

  68. Re:The Truth is by roadrunnerro · · Score: 1

    Finest carrier? The same that censors internet access for large swaths of a continent based on moral concerns of some organisation in one country? (and I'm not talking about China - it's Europe and Germany)

  69. Point taken. Thanks. [nt] by Luyseyal · · Score: 1

    Point taken. Thanks.
    -l

    --
    Help cure AIDS, cancer, and more. Donate your unused computer time to worldcommunitygrid.org. Join Team Slashdot!
  70. Baloney. Its just bad companies. by Spazmania · · Score: 2, Informative

    If either Level3 or Cogent was buying a "default" service from a third party, their customers wouldn't have a problem. The moment the peering connection was cut the lower-priority BGP routes from the third party would have taken over and their traffic would have gone through the third-party link.

    The reason these two jokers are having this problem is that they made a business decision to only move traffic with reciprocal peering and then failed to keep that peering alive. That's because they're both cheap-ass bastards; peering costs a heck of a lot less than buying transit.

    Go buy from someone else who who isn't a cheap-ass. Someone who buys transit for anything they can't peer. You won't have a problem.

    The only lesson here is that most time honored of lessons: you get what you pay for.

    --
    Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
  71. Not true by geekoid · · Score: 1

    There is a huge need to mandate.
    Of course that need is generated by cmpanies that want the internet 'under control'.

    So if you sit back and 'let the market decide' those who controll the mark will. It won't be for you.

    --
    The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
  72. Re:The small should pay for the big? (mod this up) by monkeydo · · Score: 2, Insightful

    The pitch is even better now. If you are an L3 transit customer, Cogent will give you free service for a year. For L3's current customers this solves the immediate problem, and they wind up multi-homed, so they don't get bitten by this in the future.

    --
    Si vis pacem, para bellum
    The only thing more annoying than a Libertarian is an (un|mis)informed Libertarian
  73. Peering is optional. Routing is mandatory. by Animats · · Score: 1

    If Level 3 doesn't want to peer with Cogent, that's fine. If they export incorrect routing information to gain a business advantage, that's fraud.

  74. It has probably been said... by cr0sh · · Score: 2, Informative
    I am late to this thread, and it has probably been said anyhow, but I want to reiterate:

    The problem isn't soley with the business arrangements between the "big providers" - oh, certain, that does have impact, but the internet would be as robust as ever, if every participant on it could be a peer.

    This is how the network was meant to be, a mesh comprised of stupid interconnects and smart nodes. Every node on the internet, from the largest colo to the smallest wireless handheld, should have the ability to be a true peer on the internet. In practice, this isn't really possible, but imagine a mesh network with a distributed p2p DNS system which many people could run if they wanted to - if only a fraction were running it, and were distributed enough, such outages might not occur (the traffic could continue to be routed, albeit at a slower pace).

    Everyone should be able to be a peer on the network, everyone should be able to get at least one static IP, everyone should be able to run their own server(s) if they want to. Right now, the only way you can do it is by paying huge amounts of $$$ so you can get a garden hose instead of a straw. I am not saying access to the internet should be or could be free, but peering should be a natural right of being a part of the internet, not something you have to pay extra (a LOT extra) for.

    --
    Reason is the Path to God - Anon
    1. Re:It has probably been said... by bonehead · · Score: 1

      I don't think you understand what the term "peering" means in this context. For one thing, this type of peering IS free, by definition. When Cogent and L3 peer with each other, no money changes hands between them. (Large sums of money are spent, however, on the physical links that the traffic passes through.)

      You have the same "right" as any of the Tier 1 ISPs to peer with any of the others. What I suspect you do NOT have is anything to bring to the table to make peering with you an attractive option. If Level 3 agrees to carry traffic from your home to their customers for no charge, what are you offering in return? What network resources do you have to offer them for free that they might find useful? Free transit over your home network to the little blog server in your basement isn't going to cut it.

  75. Roadrunner affected by Jeff85 · · Score: 2, Informative

    I had a friend on Roadrunner who complained he couldn't connect to many sites. I think he happened to know that they used Level 3. Is there a way to determine what backbone your ISP or a particular site uses?

    --
    Fetch Text URL - Firefox Extension
    1. Re:Roadrunner affected by Secrity · · Score: 2, Informative

      Is there a way to determine what backbone your ISP or a particular site uses?

      The Unix traceroute command can be used to do this:

      $ traceroute slashdot.org

  76. It's Nobody's Fault by quibbs0 · · Score: 1
    Everybody wants to point the finger somewhere. The fact of the matter is, the link between Cogent and Level 3 was a major one. Many people took it for granted and didn't need an alternate route. Now they do.

    I can see both sides of the argument. Level 3 says they are losing money and Cogent says "hey we are sharing! WTF".

    This has got stuff screwed up pretty bad though. The mesh topology I thought that existed apparently does not. I'm assuming DNS is all whacked out too. My company belongs on the Cogent side of this and our bank happens to be on Level 3's side (in networking terms). So, we are unable to do any banking (hourly and daily wire transfers, etc). I assume if the bank was multihomed this wouldn't be a problem but it looks like there is only one way in for now.

    The oddest part...I did a WHOIS at networksolutions.com for our banks website and it says they are 1.1.1.1??? Anybody seen that kind of behavior before?

    1. Re:It's Nobody's Fault by Alioth · · Score: 3, Informative

      Cogent COULD route around the damage - if they wanted to, but they don't.

      If the peering point had been taken out by a bomb, the re-routing would have been performed in fairly short order. However, this is not the case here.

      Level3 think that Cogent is taking the piss and is not a real peer. Level3 want Cogent to buy transit to reach Level3, either directly from them (or from someone else) because at the moment the peering is very lopsided, and costing Level3 a bucketload of money and giving Cogent a boatload of free bandwidth.

      Cogent on the other hand doesn't want to pay for transit to Level3.

      Right now, Cogent could route all their traffic for Level3 over transit they pay for. They don't want to do that because it won't force Level3 back into the peering agreement. So what they do is leave the link severed and do not re-route so that Level3 customers cannot get to sites hosted by Cogent. This means Level3 customers will grumble at Level3. Additionally, they offer a year's free transit to single homed Level3 customers just to raise the brinkmanship with Level3 a notch higher. Basically it's war between L3 and Cogent.

      If Cogent re-routes their traffic, they are defeated and L3 will never re-peer. What Cogent are hoping is that enough angry customers on the L3 end will whine at L3 so L3 will be forced to re-peer.

      For the rest of us in the peanut gallery (i.e. those of us who aren't single homed customers of Cogent or Level3) we can just watch the fun and games and throw peanut shells at the squabbling combatants because we don't see any black hole at all.

    2. Re:It's Nobody's Fault by quibbs0 · · Score: 1
      Quite an interesting view there. This whole thing all together is really fascinating actually. Yesterday I thought it was all a DNS problem early in the morning then I started putting the pieces together.

      I agree with you on your views there. Cogent could move on with the divorce but instead they are trying to get back together with their unreliable wife.

      It's funny here too cause I get to sit back and watch the accounting people flip out since they can't get to the bank and I get to laugh and say "THERE'S NOTHING I CAN DO" where as there normally is. This is a very unique situation and I'm definately interested to see how it all pans out. One way or the other, if we aren't back on the banks website next week I'm sure there are going to be some serious calls flying around.

      I can't imagine how this is effecting large corporations right now. Hell we are just a small/medium business. Then again, a large corporation is probably not single homed. :)

      ATLEAST I HAVE SLASHDOT! I can do without wikipedia and the bank.

    3. Re:It's Nobody's Fault by bonehead · · Score: 1

      Then again, a large corporation is probably not single homed. :)

      At least not the smart ones. :)

      Hell, I work for a fairly small company that is connected by a single T1 to a small local ISP. They are multi-homed to 3 separate Tier 2 ISPs, each of which is multi-homed to many places (including both of the players in the current pissing match). So we're pretty well protected against this particular sort of drama.

      Seems that by going with a small, local ISP we actually have better redundancy than if I had tried to be l33t and buy our T1 from Level 3. While our T1 is obviously a single point of failure, as long as that stays up, we have several layers of multi-homing between us and the Tier 1 guys. It would take some serious pissing matches between quite a few of them before I experienced an "Internet partition".

  77. Re:Net blackouts by the+arbiter · · Score: 1

    You know what? You're an asshole with no sense of humor. Think before you post, indeed.

    --
    Boycott everything - they're all trying to fuck you one way or another
  78. Re: Fucking Kids stuff by EddyPearson · · Score: 2, Insightful

    HOW PATHETIC! Two major ISP are willing to piss off thousands of people just because they've thrown their toys out of the pram. Back at school they'd get told to shut up and get along, now it'll become a legal action. GET A FUCKING GRIP!!! I thought the world was too sensible for this kind of thing. I was wrong.

    --
    You feel sleepy. Close your eyes. The opinions stated above are yours. You cannot imagine why you ever felt otherwise.
  79. Must.. not... reply... by sean.peters · · Score: 1

    Must... not..

    All your internets are belong to UN!

    Damn. Couldn't stop myself.

  80. A business problem is a technical problem. by hackwrench · · Score: 1

    Management is a technology.
    The business decisions affect the layout of the internet. The two are intertwined, and if there is an solution to this problem that involves a change in architecture, well then, there's something wrong with the current architecture.

  81. Now do you understand Google? by n9mdh · · Score: 1

    Seems that buying up dark fiber might be a very lucrative investment. Your backbone in a pissing match? Jump on GoogleBone....

  82. The most infuriating tech call I ever took by doublem · · Score: 1

    A backhoe took out Verizon's main cable into a few towns in Massachusetts. The reason this was an issue for us is it included the town that housed the hosting facility that had our servers. Our sites were down for most of a day.

    One guy called up, demanding to know why out site was done, and he kept telling me to "Reboot the file sever" and "I don't care about your server. You don't need a server to have a web site. I have a web site and I don;t have a server."

    Mind you, this jackass was a middle manager trying to take online courses, not an employee.

    I went through the spiel that I'd been giving all the people who called up. The typical reply was along the lines of "SO it should be up tomorrow" or "Bad luck." A few people said "Don't you have another data center?"

    This jackass said, "The Internet can't go DOWN," in the tone of voice normally used by high school students trying to humiliate someone with supposedly superior knowledge.

    "Sir, I never said the whole Intern-"

    I then heard him yelling to other people in his office, "The a**holes say the whole Internet's down. I can't believe they expect me to believe this kind of-"

    Someone else took the phone and said to me, "The INTERNET isn't down, just your piss poor site."

    "Sir, I never said the Internet was do-"

    This new person yelled t the rest of his office, "They're changing their story."

    Now I was pissed.

    "Listen, I am NOT changing my story. The building that has our servers lost it's Internet connection. A cable was cut by a jackass laying cable. When they fix the cable, we will be back online. I never said the Internet was down."

    "That's not what Bob said you said."

    "In which case I question his listening skills and comprehension of technology." What the Hell, I was already job hunting.

    Dead silence at the other end.

    "Give me your manager."

    "I AM the manger," I replied. The only person above me is the company's owner.

    "Give me the owner."

    "He left for the day when he found out our hosting center had lost it's Internet connection. I can send you to his voice mail if you like."

    "Give me someone in sal-"

    "They left too."

    "Give me the owner' voice mail."

    I did.

    In the end, the owner had so many angry voice mail messages about the outage that he went through and deleted them.

    He was very happy when I reminded him and the sales staff that we were still at 99.9% up time.

    --
    "Live Free or Die." Don't like it? Then keep out of the USA
  83. Why don't terrorists target undersea cables? by funtime · · Score: 0

    I was just wondering, if an outage can cause such trouble, why don't antisocial elements such as terrorists target intercontinental cables and suchlike? Of course, they won't have the resources to hit them on the seabed, but couldn't they get them where they enter the water?

  84. Not really.... by Rolan · · Score: 1

    The reason that Cogent is still blacked out from Level 3 customers is that the traffic routing's not being changed. There's still connections between the two networks (through other networks), but those aren't being used by choice. Cogent COULD fix it if they wanted to, but they rather have the blackout and try to blame Level 3 for it.

    --
    - AMW
    1. Re:Not really.... by lostlogic · · Score: 1

      http://ask.slashdot.org/comments.pl?sid=164413&cid =13726522 According to this post, it's not cogent that is blocking, it's L3.

      --
      --Brandon
    2. Re:Not really.... by bonehead · · Score: 1

      Cogent COULD fix it if they wanted to, but they rather have the blackout and try to blame Level 3 for it.

      Well, not exactly.

      Yes, Cogent could fix the problem if they wanted to, by routing through other carriers that they purchase transit from. Obviously, they want to avoid this because it would cost them a shitload more than the free peering arrangement with Level 3. To make matters worse, from a political standpoint, once they start routing this traffic over a paid route, their chances of getting Level 3 to reinstate the peering drop from "slim" to "none".

      In the end Cogent will quite likely end up routing the traffic over paid-for transit, but not until they've exhausted every option they can think of for getting Level 3 to reinstate the peering.

    3. Re:Not really.... by cswiger2005 · · Score: 1

      It wasn't Cogent's routers which were returning ICMP host unreachable messages; Level3 decided to blackhole the Cogent netblocks. As I just mentioned elsewere in this thread, there isn't a problem if the two depeered, and decided to route traffic between them via some third party. However, when the Level3 routers claim that anything at Cogent is unreachable, it's Level3's responsibility and fault to correct.

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
  85. This article confuses and infuriates me! by BELG · · Score: 1
    In theory, this kind of blackout is precisely the kind of problem the Internet was designed to withstand.


    Yeah, it was designed so that when two parties don't want to exchange traffic anymore, magical InterWebGnomes charge in, armed to their teeth, to correct the problem and dicipline the offenders. In fact, the very ability to filter traffic at all means the network is broken! Oh, the horror!
  86. So maybe we need the EU to intervene here by bobbo69 · · Score: 1

    ;) Seriously though - more regulatory intervention needed perhaps?

  87. But what are their terms of service? by Ungrounded+Lightning · · Score: 1

    OK, so Cogent and Level 3 don't want to talk. Fine. That's their business agreement between each other.

    But what about the business agreements between them and their customers?

    Have they violated their contracts with their customers? Do they owe them a refund for the full time the Cogent/Level 3 link is down? That could be a lot more expensive to both than accepting a more painful peering agreement, or buying transit from a third party for the duration.

    Have they written the TOS so that they can do this without violating it? Then the customers are SOL. But in that case the customers have a strong incentive to hunt up another provider with better contract terms. So this could lead to a major, and ongoing, erosion of the customer base for both companies.

    I suspect the invisible hand is about to give them both a MAJOR spanking. And give the net a push toward avoiding recurrences of this sort of "failure".

    Internet service customers aren't paying to be pawns in this game. They're paying for connection to the internet - ALL of it. To the extent that Cogent and/or Level 3 convinced their customers that internet service was what they were buying, if they entered the contract while willing to take this move they engagaged in fraud.

    Smart move for whomever DIDN'T initiate the line cut would be to cut a deal with a third party to pay them to carry their packets and deliver them TO the other. (The other can follow suit or not as it choses - but the willing player can't suck out packets the unwilling won't route.)

    Then the willing player can say to its customers "I'm keeping my contract with you. I'm getting your packets delivered to their routers. Whether they go the rest of the way or not is the other guys operation. I'm accepting packets for you from his customers wherever they hit my routers - if they ever do - and delivering them to you. If the other guys won't deliver them to me there's nothing more I can do technically. I've tried to maintain a direct connection with him but he won't budge on reasonable contract terms so I'm paying extra to do everything I can without that to give you the service I sold you."

    A solution of that form would keep a single bully from taking down anything but his own customers - and do it in a way that would be visible to any customers of the company he's fuding with: A traceroute would show that their packets go to the third party and stop at their peering router with the kermudgeon (which will probably be named appropriately) the first kermudgeon router, further in where filtering takes place if inbound packets are cut off, or all the way to the target. Meanwhile a traceroute from the target would show the packets never leave the kermudgeon's net. B-)

    Of course if the kermudgeon decides to pay for the third party transport to keep from looking bad, the net is back up, while both sides have an extra cost of business that gives them an incentive to cut a cheaper deal.

    --
    Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
    1. Re:But what are their terms of service? by cswiger2005 · · Score: 1

      My company has been affected by this pretty signicantly, although our regional ISP (reynwood.com) has done a fair amount to set up alternate routes for us since Level3 started blackholing PSI.net/Cogent netblocks. The problem is not that Level3 and Cogent depeered, but that Level3 routers were blackholing the Cogent netblocks by returning ICMP host unreachable messages. Once Level3 decided to turn off their direct connection with Cogent, then they should have chosen to route the traffic via one of their other peers as an intermediate, not claimed that all of Cogent's network was unreachable.

      Here's a bunch of traceroutes to some of my Usenet peers showing exactly what was going on:

      % traceroute news-in.newsgroups.com
      traceroute to news-in.newsgroups.com (38.119.100.108), 64 hops max, 44 byte packets
        1 polycom1.codefab.com (199.103.21.254) 0.947 ms 0.787 ms 0.874 ms
        2 199.103.21.9 (199.103.21.9) 3.011 ms 3.378 ms 3.225 ms
        3 sw1.32a.nyc.reynwood.com (199.103.19.125) 3.560 ms 4.381 ms 4.361 ms
        4 ge-8-1-241.core1.NewYork1.Level3.net (4.78.160.45) 3.960 ms !H * 4.476 ms
      !H

      % traceroute news-in.spamkiller.net
      traceroute to news-in.spamkiller.net (38.119.71.4), 64 hops max, 44 byte packets
        1 polycom1.codefab.com (199.103.21.254) 0.946 ms 0.771 ms 0.859 ms
        2 199.103.21.9 (199.103.21.9) 3.081 ms 3.457 ms 3.312 ms
        3 sw1.32a.nyc.reynwood.com (199.103.19.125) 3.896 ms 5.366 ms 4.283 ms
        4 ge-8-1-241.core1.NewYork1.Level3.net (4.78.160.45) 3.986 ms !H * 4.455 ms
      !H

      As an update, according to this status page here at http://status.cogentco.com/, Level3 has blinked and is restoring connectivity for now while both sides negotiate:

      10/7/05 (4:20pm edt): Level 3 has restored all peers with Cogent as of 4:00pm edt. We are seeing some latency in traffic across to Level 3 as sessions re-establish, mail servers deliver messages, etc. We hope the above normal traffic volumes will decrease within the next hour.

      10/7/05 (3:40pm edt): Level 3 has restored some of their peering sessions with Cogent at this time. We do not know if this is a temporary or a permanent change, and will continue to negotiate with Level 3 to resolve all the issues they have with Cogent.

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
  88. Illegal? by Renraku · · Score: 1

    This should be illegal. When you run an ISP, you basically say, "I'll take traffic and route it." along with the usual service providing to individuals or companies. Blocking any section for business reasons is just retarded.

    What if I built a really popular road with my own labor and materials, and said, "Sure! You people can use it..no problem.."

    And then I decide that only cars I SELL can drive down that road. Everyone else has to keep off. The government would smack me down and take it from me.

    --
    Job? I don't have time to get a job! Who will sit around and bitch about being broke and unemployed then?
    1. Re:Illegal? by bonehead · · Score: 1

      If you build the road on privately owned property, with privately purchased materials and labor, there would be no problem with you permitting or denying access according to your whims.

      It should also be noted, that as of 4:20 EDT today, Level 3 has turned the peering connections to Cogent back up.

      http://status.cogentco.com/

  89. But there IS a solution by frost22 · · Score: 1
    Kidding aside: For those of you who don't quite gather the scope of this: Georgetown University and (as mentioned by the article cited above) The Museum of Fine Art in Boston are just two example of organizations *heavily* affected by this. Oh, and there's no resolution on the horizon, either.

    For those of you that haven't already done so, you might consider a trip to your local Attorneys General web site to file a formal complaint.

    But there is a perfectly viable and simple solution:

    "Dear ISP,

    as you might know we are your Customer, Contract Number .

    According to our contract you are to provide internet connectivity to us for a monthly fee of . Unfortunatley you failed to do some since Wednesday Morning this week. I assume this is because you yourself are a Level3 customer, and Level 3 can't deliver that type if service any more.

    Please correct this situation within the next 12 houzrs. Internet access is essential for your business and you are already at the borderline of your SLA. Personally I'd suggest you buy additional upstream from someone else than Level 3 (Cogent might seem a good Choce, sinmce they currently offer it for free to you) but of course the decision is yours, wahtever gets the job done.

    Please be advised, though, that we definitely expect you to restore service no later than 12 hours after you have been served this notice. Failing to to so will force us to sue you for breach of contract in court, and since you elect not to restore our service despite of beeing able to do so and despite of us having fulfilled our part of the contract, we consider you acting in bad faith and consider the damage disclaimers in our contract void.

    We look forward to your cooperation in this matter.

    Joe. Q. Customert, Inc."
    --
    ...and here I stand, with all my lore, poor fool, no wiser than before.
    1. Re:But there IS a solution by Anonymous Coward · · Score: 0

      That works... but only to a small extent. First, let's face it, it goes into the 'Trash' bin if emailed. The techs who receive these are powerless.

      Second, this does NOTHING to help the customers of ISPs who are being jerked around by Level 3's actions. The ISP paid its bills. The customer paid theirs.

      They have ZERO relationship as customers with Level 3. As far as I'm concerned L3 is 100% out-of-line in this.

  90. Cogent and Level 3 intersects are at CRITICAL by digitaldc · · Score: 1

    dereference, thanks for the link...everything is green except where these two meet

    --
    He who knows best knows how little he knows. - Thomas Jefferson
  91. Re:Net blackouts by crazygeek02 · · Score: 1

    Good grief. :-)

  92. Obligatory bash.org quote... by Anonymous Coward · · Score: 0

    ...wait, I can't seem to get on bash.org today. Wonder why.

  93. And they want to turn it over to the UN? by JhohannaVH · · Score: 1

    Ok, and peeps want to turn over the root DNS Servers from ICANN to the fUN*? WHAT?! This so reminds me of the old pissing contents back in the days over phone line bandwidth when I worked for smaller telcom companies. They would have to negotiate constantly with the big boys just to let their little traffic onto their networks. *sigh* When did things become so difficult?

    --
    Sorry man... the Internet pooped on me.
  94. Userfriendly? by wembley+fraggle · · Score: 2, Interesting

    Is this why I can't read userfriendly or Something Positive this morning? Or is it just some weird coincidental webcomic blackout?

  95. Fortunate are the ones ... by really? · · Score: 1

    ... who have kept their UUCP skills alive. For they are the only ones who will be getting their pr0n in the new age. ;-)

    --

    "Consistency is contrary to nature, contrary to life. The only completely consistent people are the dead." A. Huxley
  96. Re:No fruit fucker by Anonymous Coward · · Score: 0

    is this the reason Penny Arcade has been site-non-grata for the past few days?

    Fruit Fucker Fans wanna know!

  97. PLEASE MOD PARENT UP by Anonymous Coward · · Score: 0

    With a "+1 Concise and to the Point"

  98. Cogent champion at ignoring deadlines by alephnull42 · · Score: 1

    Cogent went through exactly the same exercise with France Telecom in April this year - I was on the receiving end (hosting a service behind for lots of french customers) and it was a right royal PITA.

    Interestingly enough, it was the same story, France Telecom had given Cogent 60 days notice to sign a new agreement. The deadline went by, Cogent wouldn't budge, never the faintest warning to the customers that there was a potential problem. In fact it was 12 hours after the peering went down before we got ANY information from Cogent, which was the following mail, basically a royal "passing of the buck".

    From: NOC-EU
    To: Thu, 14 Apr 2005 10:06:13 -0400
    Subjact: Peer Problem with France Telecom HD271257

    Dear Cogent Customer,
      We are very sorry to hear you are experiencing problems with your
    connection. We're aware of this issue and we are working to improve
    things as quickly as possible.

      The specific issue is with France Telecom. Since Cogent's inception,
    we have maintained a robust peering relationship with them. Recently,
    however, they made a unilateral decision to de-peer. The direct
    result
    of this action is the isolation of our two networks.

      Cogent is very concerned by this action. Peering relationships such
    as these are the foundation of a successful Internet community. We
    are
    working to better understand France Telecom's motivation. We
    speculate
    that this action is in direct response to Cogent's competitive
    pricing
    in the market. Cogent's goal is to have France Telecom re-instate the
    connection.

      We appreciate your patience as we work on this issue. We are closely
    monitoring this situation in our support center. If you call back for
    an update, please reference master ticket: HD271257 We encourage you
    to also contact France Telecom so that they fully understand the
    impact of their action. You may reach them at noc@opentransit.net

      As always, we sincerely appreciate your business. We will keep you
    up
    to date on the situation.

      Sincerely

    regards / Mit freundlichen Gruessen

    C.C.D. C o g e n t Communications Deutschland GmbH
    Stephanstr. 3
    60313 Frankfurt

    Tel.: +49 69 299896 96
    fax.: +49 69 299896 54
    mail: noc-eu@cogentco.com

    It seems to be Cogent's business philosophy to play a game "chicken" in these situations, to not warn the customers, and to take it's customers as hostages.

    We migrated our entire uplink to another provider in 3 days, and of course the link came back up on day 5 - but we're NOT going back.

    --
    Not confused enough? http://translate.google.com/translate?u=www.slashdot.jp&hl=en&ie=UTF8&sl=ja&tl=en
  99. Greed can... by Anonymous Coward · · Score: 0

    ...sabotage even the best infrastructure.

  100. not really by brunes69 · · Score: 1

    They wouldn't have been a transit until that link broke down.

    Say you have 3 huge backbone providers, A, B, and C.

    A peers with B
    B peers with C
    A peers with C

    None of these providers is a transit provider for the others, because none of them need it as they all have peering relationships.

    Now say the deal between A and C goes bad; the traffic can still get bwtween them by going through B. Now, in this situation, B would become a transit carrier, yes. I don't know if that would mean immediately that they would have to start paying higher fees or what, because that would not be their original contract

    1. Re:not really by gskouby · · Score: 1

      We are saying the exact same thing.

  101. INTERWEB KAN SURVIVE A NUCLEAR ATTACK! by Harik · · Score: 1
    ... But as a private ISP, L3 and Cogent are allowed to not accept packets from someone they're having a business dispute with. L3 feels cogent is misrepresenting their status to qualify for peering, Cogent doesn't want to pay someone for transit when they can dump the packets for free. Exacerbated by the fact that Cogent is completely undercutting L3's prices, and to L3 it feels like they're doing it by cost-shifting unto other carriers using free peering rather then paid transit.

    Fun!

    As was pointed out numerous times, both of them have many other peers in common that could correct this with a few quick router commands. It would be expensive (multi-thousands per day), but in a disaster situation it would be done. We're talking about 7.5% of L3's customer base and 11% of Cogent's base having partial internet unreachability. This isn't a widespread OMG SKY IS FALLING event. Hell, we're all still posting this on slashdot VIA THE INTERNET WHICH IS OPERATING AS DESIGNED.

  102. In Other News by jnadke · · Score: 1

    Beijing (AP) - In a move that has surprised the world, China has announched today that its new DVD format will be 100% Freedom-Free. "We want to make sure terrorists cannot attack the pride of the People's Republic of China," said President Hu Jintao. "China will not be hindered by other formats that could possibly include Freedom protocols," he concluded.

    "We were just trying to stop those damned file sharers," said Mitch Bainwol, Chairman and CEO of the RIAA. "This time, China has gone too far. They can't expect to attack freedom and get away with it. Besides, how are we supposed to be the bad guys when China shows us up with this? We have an image to maintain."

    Following the announcement, the RIAA is expected to respond later today with a Data-Free DVD format. "You can't steal what you can't see," said Bainwol.

  103. PARENT - Ignore, Wrong Story by jnadke · · Score: 1

    Didn't realize I was on a reply for a different story. Got my FireFox tabs messed up.

  104. The peering is back up by Anonymous Coward · · Score: 2, Informative

    the problem has been solved. I can ping level3 from cogent and i have one connection. I don't know yet who flinched first......

    1. Re:The peering is back up by Anonymous Coward · · Score: 0

      Maybe Level(3) responded to Cogent's press release:

      "Cogent Communications Group, Inc. (AMEX: COI), a Tier 1 Internet Service provider, is requesting that Level 3 turn their side of the companies' Internet backbone peering connection back on immediately to thwart any further disruption of mutual customers' Internet service. Once traffic is being exchanged between the two networks, Cogent will discuss the peering situation with Level 3,anytime, anywhere."

      http://www.cogentco.com/htdocs/press.php?func=deta il&person_id=62

      Cogent's looking glass also confirms this http://www.cogentco.com/htdocs/glass.php:

      trace www.level3.com

      Type escape sequence to abort.
      Tracing the route to www.level3.com (209.245.19.42)

          1 g10-0-224.core01.lax01.atlas.cogentco.com (66.250.4.5) 0 msec 4 msec 0 msec
          2 p14-0.core01.sjc01.atlas.cogentco.com (66.28.4.74) 12 msec 12 msec 12 msec
          3 p4-0.core01.sfo01.atlas.cogentco.com (66.28.4.93) 16 msec 16 msec 16 msec
          4 p15-0.core02.sfo01.atlas.cogentco.com (66.28.4.70) 12 msec 12 msec 16 msec
          5 p10-0.core01.sjc03.atlas.cogentco.com (66.28.4.133) 16 msec 16 msec 12 msec
          6 so-0-2-0.edge2.SanJose1.Level3.net (4.68.127.201) 12 msec 12 msec 12 msec
          7 so-3-2-0.bbr2.SanJose1.Level3.net (4.68.121.197) 16 msec 16 msec 16 msec
      etc...

  105. Fixed now? by dereference · · Score: 3, Informative

    The availability grid for the past 4 hours shows ~40% and the grid for the past 1 hour shows 100%. As noted by "Cally" below, I honestly have no idea how exactly this grid has been generated (hence my original disclaimer) but this certainly seems to indicate, from a practical standpoint, that the L3/Cogent issue has been very recently resolved. Indeed, from my (single-homed) L3 server I can now traceroute directly to a (single-homed) Cogent host.

  106. Just another example by Trogre · · Score: 1

    ...of how capatilism doesn't work.

    These providers have no obligation whatsoever to provide any level of redundancy so long as you keep paying their bills.

    --
    "Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
  107. That's not how peering works - here's the diff by billstewart · · Score: 3, Informative
    There are two basic ways that networks connect to each other - peering and transit. In a transit arrangement, one network (typically the big one) agrees to deliver any traffic the other network hands it, in return for a bunch of money, and it typically either advertises a default route (telling a small customer that they can send it all their packets) or a bunch of detailed routes and a default (telling a dual-homed medium-large customer how good its connections are to lots of places, but that customer might use another carrier for destinations that are closer with that carrier.) If you're an end customer, or a small ISP buying service from a big ISP, that's usually what you buy.

    Peering arrangements are different. Two networks that have a lot of traffic for each other will set up direct connections, split the direct costs of the connections, and not charge for accepting packets from the other carrier. But they'll only advertise the routes for their *own* customers. If two small ISPs peer with each other, typically they're each also buying transit service from big ISPs, but it's cheaper for them to dedicate a connection or put bits on a public peering point like MAE-West than to both pay their upstream ISPs.

    The biggest ISPs in the US are called "Tier 1" ISPs, and they all peer with each other rather than buying transit, though they might buy transit for international connections, if they can't get the other side to buy transit from them. It seems flaky, but it makes business sense, or at least it did for a while. In some sense, being big enough that all the other Tier 1s will peer with you is what defines Tier 1, and aside from technical issues, it's a marketing thing - "See, we're one of the big players!" Peering and Transit don't mix very well - you either connect to a given carrier by peering, or by transit, or else you spend a long time hammering out custom arrangements about exactly which routes you'll accept and tweaking routing tables.

    Cogent is a Wannabe-Tier-1. Their main business model is to put fiber into big multi-tenant office buildings and sell everybody 100-meg Ethernet for about the price other carriers charge for one or two T1s. If I were a customer, I wouldn't expect there to be enough upstream to really get that much bandwidth all the time, but I'd expect to get more than a T1 all the time, and a lot more than a T1 almost all the time. Level 3 has apparently decided they're not getting enough value out of the relationship (i.e. not sending Cogent enough packets to make it worth their while) to keep peering, and wants Cogent to either pay them for service or get transit from somebody else. They gave them about 50 days to make other arrangements, but Cogent decided to play chicken with them.

    --

    Bill Stewart
    New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
    1. Re:That's not how peering works - here's the diff by ePhil_One · · Score: 1
      Cogent is a Wannabe-Tier-1.

      Right, but Level 3 is?

      Their main business model is to put fiber into big multi-tenant office buildings and sell everybody 100-meg Ethernet for about the price other carriers charge for one or two T1s.

      Their business model is to buy the networks of other Tier 1's when they fail on the cheap. I suspect they are also leasing space on other provider's fiber so they don't spend millions laying even more redundant fiber. They have also adopted new, non-telco based technologies that let them get more bandwidth out of their circuits for less $$. They also have Colos and other facilities

      If I were a customer, I wouldn't expect there to be enough upstream to really get that much bandwidth all the time, but I'd expect to get more than a T1 all the time, and a lot more than a T1 almost all the time.

      And you'd be wrong. I honestly didn't expect much from them either, but we've conistently flowed 50+ Mbps though the circuit. Makes me feel like an ass for spending 12x that much for an OC-3.

      Level 3 has apparently decided they're not getting enough value out of the relationship (i.e. not sending Cogent enough packets to make it worth their while) to keep peering, and wants Cogent to either pay them for service or get transit from somebody else.

      Peering, not transit. My guess is that Level 3 is having cash flow issues, despite their claims of "record earnings", though that just may be the management trying desparately to find new sources of income to maintain those record earnings.

      From what I can glean, Level 3 is heavy in the Broadband providers (users) (one of the largest providers of wholesale dial-up service to ISPs in North America and is the primary provider of Internet connectivity for millions of broadband subscribers, through its cable and DSL partners), Cogent is heavy with the Service providers (servers). From Level 3's press release:

      For example, Cogent was sending far more traffic to the Level 3 network than Level 3 was sending to Cogent's network

      They need each other.

      They gave them about 50 days to make other arrangements, but Cogent decided to play chicken with them.

      I'd say it was the other way around. Level 3 tried to extort $$ from Cogent; Cogent didn't give in. Now Level 3 has to explain to those recipients of that traffic why they cut them off; my guess is a lot of them won't like L3's explaination of "it wasn't fair" and will sign with a new provider (maybe Cogent!) when their contracts end. The perfect irony will be if Cogent buys L3 in two years when they go bankrupt from lack of clients. I have no intention of moving away from Cogent, I believe they did the right thing.

      --
      You are in a maze of twisted little posts, all alike.
  108. It's a prisioner's dilemma by Harmonious+Botch · · Score: 1

    Peering is an iterative prisoner's dilemma. The first company to overuse the other's bandwidth - ie to defect - shows a profit for having done so. As Robert Axelrod ( 'Evolution of Cooperation', 1984 ) and others have demonstrated, the best strategy in such a situation is a tit-for-tat, that is: I defect only after you do, but cooperate otherwise. Cogent defected, and then Level 3 defected. It will cost them both, but the lesson will not be lost on others. If most companies adopt TFT, those who defect prematurely will be culled out, and the the result will be a pool of better behaved companies. The resulting evolution is better for all of us.

  109. Re:Ask Slashdot by ReverendLoki · · Score: 1

    Actually, avoiding a blackout like this one is not that difficult. All you have to do is make sure you are receiving your connection either directly from, or via a provider that receives a connection from, a level 2 service provider that has signed up with multiple level 1 providers. The only ones effected by this outage are those who only have one or the other of these two level 1s (well, maybe a 1 and a 1.5). Even those that have a different level 1 are unaffected.

    --
    09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0
  110. Business routing by sjames · · Score: 1

    The problem is really a matter of routing policy. Since the net is comprised of businesses, it all comes down to money. Technically, peers should announce all of their routes to peers. Practically, they only announce routes to their own customers to their peers. So, A, B, and C all peer. Peering between A and C goes down. However, B does not announce routes to either A or C even though they know them so traffic between A and C is cut off.

    As long as the net is run by businesses that must make money to continue, and they are not paid to announce full routing tables, the net will be subject to this sort of dane brammage.