Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Japan Networking Earth The Internet News

Net Sees Earthquake Damage, Routes Around It 177

Posted by Soulskill from the we-are-borg dept.
davidwr writes "Japanese internet outages mostly healed themselves within hours. While some cables remain out, most computers that lost connectivity have it again. From James Cowie's blog: 'The engineers who built Japan's Internet created a dense web of domestic and international connectivity that is among the richest and most diverse on earth, as befits a critical gateway for global connectivity in and out of East Asia. At this point, it looks like their work may have allowed the Internet to do what it does best: route around catastrophic damage and keep the packets flowing, despite terrible chaos and uncertainty.' Let's hear it for redundancy and good planning." Reader Spy Handler points out another article about how redundancy and good planning are preventing disaster at Japan's troubled nuclear reactors, despite media-fueled speculation and panic to the contrary.
This discussion has been archived. No new comments can be posted.

Net Sees Earthquake Damage, Routes Around It More

Net Sees Earthquake Damage, Routes Around It

Comments Filter:

  • What does this have to with the reactors? (Score:1)

    by Anonymous Coward

    Good job as always, /. editors. If you wanted another nuke article, why not just post one? :/

  • This Was The Whole Point of the Internet (Score:1)

    by Anonymous Coward

    Before it was commercialized, the whole point of the internet was to create a communications system that could survive a nuclear war. Now, for whatever reasons, most countries have singular backbones and connection, and when that one is taken out, the com system designed to survive a nuclear war can disconnect an entire country because of a single boat anchor.

    Looks like another thing that Japan took from the US, and maintains it to higher standards.

    • Looks like another thing that Japan took from the US, and maintains it to higher standards.

      So the first two E's are already done, if I were you I'd watch out for the third sometime soon.

      • You mean like it already happened in, say, car industry, computer industry, audio and video industry, ...

  • Damage has been done, hello oil and coal... (Score:4, Interesting)

    by mlts ( 1038732 ) * on Monday March 14, 2011 @05:35PM (#35485346)

    Even though the Japanese reactors did their job to contain against a meltdown, it looks like nuclear power progress will be set back another 20-30 years due to the fearmongers pointing to this.

    The loss of life can't be ignored. For people that were not affected by loved ones killed by it, the rest of the world will also be feeling this disaster in Japan for generations to come. Especially the fact that the anti-nuke crowd now possesses another "kill point" to keep nuclear power dead. This essentially clinches the fact that our kids and grandkids will still be having their lights powered by coal, and their cars by oil.

    • Re: (Score:1)

      by spud603 ( 832173 )
      What about all the other power sources? It's not simply between petro and nuclear. Even if this kills nuclear there's still hydro, solar, wind, biofuels and good old conservation.

      • Re: (Score:2, Insightful)

        by Nadaka ( 224565 )

        hydro is one of the most environmentally destructive forms of power, with burning forests being worse. It utterly devastates river ecology, floods vast tracts of otherwise useful and fertile land and is currently leading to the extinction of most of the planets major migratory fresh water fish.

        Biofuel is one of the most socially destructive forms of power. Just to replace the US motor vehicle transportation costs, you would need to sacrifice nearly 100% of our food producing farmland. Note that the US provi

        • Re: (Score:2)

          by MavenW ( 839198 )

          hydro is one of the most environmentally destructive forms of power, with burning forests being worse. It utterly devastates river ecology, floods vast tracts of otherwise useful and fertile land and is currently leading to the extinction of most of the planets major migratory fresh water fish.

          Not necessarily. Only if there is a big dam with a reservoir behind it. Hydro can be done without the dam, and it's just as efficient. It doesn't have the bonus of evening out the annual flow fluctuations, but it solves the flooding and migratory fish issues.

          Close to where I grew up there was a small hydroelectric power plant of this type. Some water was diverted into a pipeline a few miles upstream. The pipe roughly followed the bank of the river, and the water gushed back into the river after turning

          • Re: (Score:2)

            by Thing 1 ( 178996 )
            Agreed. I just did a thought experiment with putting lots of little waterwheels on Niagara Falls. I searched and see that they're already doing it [niagarafrontier.com] (well, using the river for power, that is; I didn't see whether there were any actually in line of the falling water, which I think would make sense, suck that gravity well as much as possible!). Still, I think a lot of power could be generated with far less environmental impact than, say, Boulder Dam et al. We just need to mine the waterfalls!

            • You'd have a hard time putting them on the falls itself because it recedes 1' / year. Your little waterwheels would have to be similarly mobile. There are things we can do to reduce that a bit but it's still a barrier to any major permanent undertaking.

              Plus, it's kind of a tourist attraction without the waterwheels. The power plants are a fair ways upstream, and because of the tourism aspect they don't run at the full possible capacity to maintain the impressiveness of the falls.

              • Re: (Score:2)

                by Thing 1 ( 178996 )
                Interesting. So, tourism dollars versus electricity dollars. I wonder if something could be designed at the bottom that absorbs the load of the falling water without turning a wheel, like with springs. (Yeah, the idea just germinated, hasn't taken root.)

            • The niagara falls power system provides a significant portion of Canada's power, and quite a bit on the american side as well. Treaties control the amount of water diverted, but they do make full use of the gravity well. Not with inefficient and awkward water wheels, but with well designed slopes leading to properly coupled turbines.

              Also, they divert the water upstream of the falls, and release it downstream, so they've actually got even more gravitational potential to take advantage of.

              And all of it is a

          • St. Jacobs near Waterloo has something like this, but on a smaller scale for old a mill. There weren't any good drops anywhere for the water wheel so they split the river and diked it around the town so when they met up again the drop was more reasonable. It doesn't wreck the environment at all from what I can tell and rivers don't stop flowing. It is a technology that has been applied for a long time.

            • Re: (Score:2)

              by stiggle ( 649614 )

              Waterwheel technology has been around for a few thousand years :-)
              The additional leat (mill stream leading to the mill from the river) and mill ponds actually provide additional habitat to the fish.

              Although in the UK - you need an extraction license to divert the river water down the mill stream, over the wheel and back into the river.

          • Hydro can be done without the dam, and it's just as efficient.

            The problem is that this isn't the case. Those dams aren't there only for the flow fluctuations, more importantly is that a large body of water results in water flowing faster through the turbines. The potential energy is much higher when you have all that pressure behind the dam.

            • Expanding on my own post, I recently looked into run-of-the-river small scale hydro but the calculations show that you'll need a fast flowing river to get any serious amount of energy. With wind-power it's easy to scale: make it higher and the blades longer. With hydro-without-a-dam you're stuck with sucking the same amount of energy from a single stream. It works for small scale power generation but it's not something that will contribute beyond that.

              For serious large-scale hydro energy without dams, the o

        • "wind is unreliable with bursts capable of damaging power transmission and occasional lulls that cover vast regions at a time."

          Wind can be used in large offshore chains with high guarantees of predictable amounts of energy. Wind energy can also be used to produce hydrogen (nd hydrogen can in turn be used to produce carbonaceous liquid fuels using CO2 from the air). Wind energy can also be used intermittenlty to crush rock for fertilizer (see remineralize.org). Wind energy can also be used to compress air in

        • And what of geothermal energy?

          • Re: (Score:2)

            by lgw ( 121541 )

            Geothermal doesn't scale. IIRC, geothermal power per square meter is about 1/10000th of solar power (averaged over the surface of the Earth, ovbiously it's not evenly distributed in practice). Nice in a few areas where it's easy, but no substitute ofr base load. Most "alternative" energy sources trip over the same hurdle: 1TW is a lot of power.

            Solar scales very well (and you don't need rare high-efficiency photoelectric cells, just a black pipe and a mirrored trench), but isn't reliable. Until we get a

        • hydro is one of the most environmentally destructive forms of power

          Really? It was always the best in SimCity 2000.

    • Re: (Score:1)

      by arunce ( 1934350 )
      That's right. It happened before and will happen again. And the lies are more or less the same.
    • The Hindenburg feels your pain. The world can adapt and continue to progress, even if we choose to kill random technologies.

      • This is different. Nuclear energy is a useful tool, the air ships in contrast were already in the process of being replaced at that point. The main reason why they were being used is that planes hadn't yet gotten to the point where they could reliably cross an ocean, let alone with enough passengers to make it worthwhile.

        I wasn't aware that Japan had nuclear reactors, it was a really dumb idea for them to do. In the US the few nuclear reactors we have are designed so that if power is lost to the core the co

        • Re: (Score:2)

          by lgw ( 121541 )

          Keeping the reactor usable after an earthquake is a much more expensive prospect than merely preventing an uncontrolled reaction leading to too much waste heat. There are several modern fail-safe ideas, from pebble-bed reactors where nuggets of fissile material are encased in a coating with a temperature-sensitive neutron cross-section (so if it gets hot, the reaction is stopped), to simple meltable housings that flood the reactor chamber with a neutron absorber if a temperature threshhol is passed (think

        • I wasn't aware that Japan had nuclear reactors, it was a really dumb idea for them to do. In the US the few nuclear reactors we have are designed so that if power is lost to the core the control rods fall into the core and the fuel rods fall out and the reaction stops. The problem is that if a reactor like that suffers and earthquake you can end up in a position where the rods get jammed and the assurance of an automatic shutdown disappears.

          From what I've gathered it's a bit of a moot point as these reactors were apparently built upside down such that they have to have constant power to keep the reactor offline.

          Note that the Japanese reactors at Fukushima which are currently melting down are a U.S. (General Electric) design [wikipedia.org], and the oldest (1 and 2) were actually built by GE... This design is apparently quite common in the U.S. as well.

    • Yes people don't understand nuclear power generation != nuclear weapons
    • How the hell is this off-topic in a conversation about the perceived safety of nuclear power?

  • anthropomorphizing (Score:3)

    by spud603 ( 832173 ) on Monday March 14, 2011 @05:41PM (#35485400)
    "Net Sees Earthquake Damage"; "[internet] routes around it"; "outages mostly healed themselves"
    Why do we insist on speaking of the internet as some mythical being with the ability to observe, act and heal? It's true that there is a remarkable robustness to the network, as shown in this case, but why do we need to attribute it to anything beyond simple 'redundancy and good planning'? It's a network of electronics and fiber-optics, maintained by people --- infrastructure and connections.
    The internet doesn't 'see' anything, and information doesn't 'want' anything.

    • Why do we insist on speaking of the internet as some mythical being with the ability to observe, act and heal?

      Its sounds nicer and we are just thinking of Skynet and the Matrix. We are simply the worker cells. :)

      • Re: (Score:2)

        by spud603 ( 832173 )
        I guess that's a good point. How are we going to organize a resistance against it if we can't say that it's plotting our demise?

    • Re:anthropomorphizing (Score:5, Insightful)

      by mcavic ( 2007672 ) on Monday March 14, 2011 @05:51PM (#35485494)

      The internet doesn't 'see' anything

      Routers do. They can see a loss of connectivity and alter their routes accordingly.

      why do we need to attribute it to anything beyond simple 'redundancy and good planning'?

      A redundant route doesn't do any good without the intelligence (either human or machine) to determine which routes are up and send traffic through them only.

      • Re: (Score:2)

        by spud603 ( 832173 )

        A redundant route doesn't do any good without the intelligence (either human or machine) to determine which routes are up and send traffic through them only.

        I guess tend to think of the software that makes redundancy act like redundancy as just a component of the redundancy (nobody says they have redundancy to their data just because they installed a second drive---the redundancy comes from the RAID implementation).
        But you're right, there needs to be some algorithmic intelligence there. But the language used in the headline here almost suggests some sort of hyper-consciousness to the internet. I think it's a lot more amazing to think of it as a crazy-complex

        • Re: (Score:1)

          by Anonymous Coward

          I think it's a lot more amazing to think of it as a crazy-complex system of interdependent parts than as some unified being.

          Strange, I think of the human brain as a crazy-complex system of interdependent parts. But then again, I do not consider all humans to be intelligent.

      • Re: (Score:2)

        by msauve ( 701917 )
        "Routers do. They can see a loss of connectivity and alter their routes accordingly."

        It's anthropomorphizing to say "they see" a loss of connectivity. Routers don't have eyes or cameras. They do have mechanisms to detect the loss of connectivity, though.

        "A redundant route doesn't do any good without the intelligence..."

        Again, it's anthropomorphizing to attribute a change in routes to "intelligence." Routers aren't intelligent - they blindly follow a well defined set of rules. Order all routes by their de

    • Re: (Score:2, Informative)

      by Anonymous Coward

      It's a modification of a famous quote by John Gilmore [wikipedia.org]: "The Net interprets censorship as damage and routes around it."

      Know your Internet heroes. (Everybody recognizes that Zuckerberg twerp, but the net was fostered by people with beards!)

    • Its not a matter of considering the net a live entity, but it is a complex mesh of devices, each of which has a specialised function and the sum of those devices makes information flow based on certain decisions.

      As large amount of decision making on routing, load-balancing, reflowing and path finding is automated and based on certain stimuli (broken links, bandwidth thresholds, lack of net neutrality, etc.) then the system in question -the internet- exhibits a behaviour which is dependant on those stimuli.

    • Because what is being described is an automatic function, carried out autonomously by the infrastructure of the net? We've programmed and built it to react to input in a certain way. In short: in sees damage to its infrastructure, it automatically routes around it and thus it "heals" that damage.

      It has the ability to observe, because we've given it ways to receive input. It has the ability to act, because we've programmed it to do so. Being able to 'heal' is a consequence of this instance of acting. It is d

  • According to AFP (Score:4, Funny)

    by ackthpt ( 218170 ) on Monday March 14, 2011 @05:41PM (#35485404) Homepage Journal

    Network traffic has moved 8 feet to the east.

    • Re: (Score:2)

      by chill ( 34294 )

      Hmmm...lower ping to servers in Japan! Woot!

    • >>Network traffic has moved 8 feet to the east.

      On a serious note, I wonder what this will do to GPS navigation systems.

      • >>Network traffic has moved 8 feet to the east.

        On a serious note, I wonder what this will do to GPS navigation systems.

        I suspect everything in Japan will be off by 8 feet or so, until new maps can be distributed.

    • Network traffic has moved 8 feet to the east.

      That was horrible. Funny - but horrible.

  • I don't know about Mostly. Getting to websites outside of Korea has been a very slow and arduous process since the quake hit. It's 3 days in and a good number of sites are still crawling.

  • Does anyone know how other countries compare in this regards? I imagine certain countries have certain clear points of failure.

    • I have a little personal experience in Hong Kong, as I live there.

      A few years ago there was the Taiwan quake, that basically severed all cables between Hong Kong and Taiwan, and effectively all but cutting us off from the rest of the world. Local and mainland China based sites worked fine; the rest was cut off for a few days. That included e-mail of course. Not really a "single" point of failure as there were multiple cables involved, still "single" enough to fail. Pretty bad. Luckily within a few days eme

  • It's not just the network architecture (Score:5, Informative)

    by antifoidulus ( 807088 ) on Monday March 14, 2011 @05:53PM (#35485526) Homepage Journal
    The network architecture isn't the only reason why we are still able to *mostly* communicate(I live about 60 km north of Tokyo, still no water though they haven't implemented the rolling blackouts....yet...), the advances in distributed systems also have made a huge impact. Simply put the amount of information to is essentially automatically mirrored(it's not really mirrored, but its easier to think of it like that) in Japan has really cut down on the amount of bandwidth necessary to communicate with the outside world.

    I have noticed that for things that almost certainly aren't mirrored and require a direct connection to the US the bandwidth is probably 1/10 of what it usually is. While some of that may be due to increased traffic, I cannot help but think given the location of the quake that some of the cables between the US and Japan have been damaged. However services like Facebook and Google are as fast as they ever were. The reason for this is simple, both Google and Facebook have data centers in Japan that are designed to be eventually consistent. Instead of each individual request being routed to the states and back almost all the requests are routed to local data centers with only the updates coming from elsewhere being pushed through the cables. This obviously saves tons of bandwidth and allows for much better communication with the outside world. Now if you'll excuse me I gotta throw out most of my stuff and get the hell out of here. Tata!

    • Re: (Score:2)

      by jrumney ( 197329 )
      News reports say a couple of undersea cables between Japan and China are out, but nothing on the US side. Most of your international bandwidth problems are probably caused by the upsurge in people watching NHK online.

      • Re:It's not just the network architecture (Score:5, Informative)

        by Anonymous Coward on Monday March 14, 2011 @08:55PM (#35487044)

        News reports say a couple of undersea cables between Japan and China are out, but nothing on the US side. Most of your international bandwidth problems are probably caused by the upsurge in people watching NHK online.

        For all of the people who are wondering about this, I am here from the source...

        I work for Global Crossing and we own the 20Gbps links that were damaged from the earthquakes and subsequent tsunami. We have ships going out today or tomorrow to lift the cables and repair them. Our reroutes are what are keeping the Internet going. For the companies reading this and who are constantly calling in, no, your connection is not worth more than human life. Reliance Globalcom, Comcast, Syfi...you are the heavy offenders. Please, it is being fixed soon as Japan allows transport in and out of the country. Please be patient and stop being brutish dicks on the phone.

    • Certainly the cable from Telia's US backbone to Asia is broken. That one *didn't* change it's route for at least the first several hours.

  • Let's hear it for redundancy and good planning. (Score:5, Funny)

    by gratuitous_arp ( 1650741 ) on Monday March 14, 2011 @06:43PM (#35485958)

    Let's hear it for redundancy and good planning.

    Let's hear it for redundancy and good planning.

    • Let's hear it for redundancy and good planning.

      Let's hear it for redundancy and good planning.

      This redundant redundancy event is brought to you by the department of redundancy department, and is organised by Chaos, Inc.

  • I don't watch the more hyperbolic networks but from what I've seen and read so far, all the statements seem to be along the lines that the threat of a dangerous radioactive leak is fairly small. However it ain't what you say it's the way that you say it. The tone of some headlines would make you think the world was about to blow up. Channel 4 News (UK) which is renowned for good quality reporting even succumbed to it in their headlines at the weekend, referring to a "nuclear emergency" which has a nice dram

    • Re: (Score:2)

      by murdocj ( 543661 )
      As usual, the truth is somewhere in between the extremes. The reactors aren't going to have nuclear explosions... but the whole "nothing bad could happen, we've got a containment vessel around it" is bullshit. If nothing bad couldn't happen, the Japanese wouldn't be working like beavers pumping seawater into the reactor. Pumping seawater means they've completely written off the reactor. It's the "nuke the site from orbit" approach. You don't do that unless the situation is pretty dire. So yeah, thinki

      • It COULD of course also mean that they don't follow the "let's try to save the plant even if there's a chance it blows" approach that I'd expect from our nuke reactor owners. I kinda prefer their "let's play it safe, I mean, we've seen what a nuke can do. Twice." approach.

      • Re: (Score:2)

        by Rhys ( 96510 )

        The thinking has already mostly been done. The problem is the sheep are too paniced for anyone to manage to build one, at least in the US. When you can build a reactor that is fail-safe* rather than fail-deadly but aren't because "nukes are dangerous!!!!" frankly we as a species deserve what we get, even if it means writing off all higher-order life down the road.

        On the flip side, China will probably build them if we don't, though I'm less sure about their real goals of fail-safe given how little they care

  • Ugh (Score:2, Insightful)

    by clyde_cadiddlehopper ( 1052112 )
    My heart sunk when I clicked on the second link ... it lead to a junk engineering article in the Wall Street Journal. Where would I go for an unbiased engineering assessment of "redundancy and good planning"? Technology Review, New Scientist, even Wired ... anywhere but the homepage of Rupert Murdoch's cadre of shills for corporate interests. He makes such brilliant observations as "water doesn't burn." No, it evaportates. Next, it dissociates in the presence of heat and certain catalysts like the zirc
    • There is a very clear, well-written article explaining about why we shouldn't be worried about the Fukushima Reactors. I live about 150km from the plant and have grown tired of the fear-mongering I see in most of the media back home. The article can be found here [mitnse.com].

      • Re: (Score:2)

        by Jorrit ( 19549 )

        Shouldn't be worried? Ok, they are not going to explode but to me any release of radioactive material in the atmosphere is something to worry about strongly.

    • Re: (Score:2)

      by WoOS ( 28173 )

      What he was saying is that you need the temperature and ashes of a "real" fire (and not just a comparably "small" hydrogen explosion) to carry nuclear waste around the world. Chernobyl will not be easily matched without something to burn, preferably something very close to the nuclear fuel.

    • Re: (Score:2)

      by quax ( 19371 )

      A much more insightful analysis can be found here [allthingsnuclear.org].

  • Well, despite optimism about how the Japan officials are handling the failures at the reactors, it seems a 3rd and more serious blast occurred in reactor II less than an hour from now... It seems melting down is ongoing. The issue here is that many things were overlooked, even if we take into account the huge magnitude of the event. For instance, the massive anti-tsunami barriers in Japan coast were no effective at all. Also, it seems many people didn't took the tsunami warning seriously and didn't go to h

  • We have a customer in Japan operating a data center in Tochigi Prefecture, only about 200 km or so from Sendai. They lost power after the earthquake, and were running off UPS until their data center gensets kicked in, so their servers did not experience any outage immediately after the earthquake. Our people on the scene reported that television and radio were out, and their only source of news was from the Internet: their connectivity seemed almost entirely unaffected. However, their generators only had enough gas for six hours of operation, so we still had to shut everything down before the juice ran out, and there was no power for eight more hours after that... I was surprised that there was no serious network service interruption: no major undersea cables were damaged like what happened after the earthquake in Taiwan in 2006, and their network performance seems just as it normally is: they still seem to be getting their advertised gigabit speed, at least to other sites also in Japan, so it seems that their net backbone was scarcely affected.

    We'll have problems maintaining service uptime in the face of the rolling blackouts that they're experiencing, but those are the breaks...

  • "At this point, it looks like their work may have allowed the Internet to do what it does best: route around catastrophic damage and keep the packets flowing, despite terrible chaos and uncertainty.' Let's hear it for redundancy and good planning"

    Yes, let's hear it for the guys that designed ARPANet...for *specifically* this kind of catastrophic, multi-node failure.

    After all, that's precisely the sort of thing the internet was, in fact, designed to cope with. So yes, let's applaud Japan for building a robu

  • Comment removed based on user account deletion
  • The internet doesn't heal itself. It's healed by on call engineers who get woken by a phone call at 3am. The automagical healing is human technical support performing heroic emergency fixes 24/7. I sure wish it could all just fix itself ...

Slashdot Top Deals

No spitting on the Bus! Thank you, The Mgt.

Close