Slashdot Mirror

Rebuild the internet (Internet II) from the hardware up, this time do it right, don't just patch it.

Internet 2 already exists.

The most likely result of a rebuilt Internet III with full security is that you won't be able to use it because your access will lessen security. It will be the lesser-used cousin to Internet I, just as Usenet II is the lesser-used cousin of Usenet.

It is still around, and doing what it was meant to do.

Here is the NOC: https://noc.net.internet2.edu/

Remember that the main reason that Internet 2 exists is that educational institions needed shelter from the Eternal September, along with all of the other crap that came when AOL joined the Internet that they and ARPA built.

This time, they kept it on-mission, and it has worked out just fine.

Very nice, but hardly new. Both ESnet (U.S. DOE research network) and Internet2, the national collegiate research network have been running at Nx100G to major research sites and the rest of the Internet for at least two years. They provide Internet service places like CalTech, MIT, the University of Califorrnia, Berkeley Lab and Fermilab. These are full production networks with ESnet already moving vast amounts of data from the LHC to the US for storage and dissemination to many public and private research facilities.

Internet2 Announces First Full-Production Virtual Internet Network Architecture
http://www.internet2.edu/news/detail/7257/

It gives me a better sense of what they're doing, but I'd still be happy to have someone dumb this down into an automotive analogy.

It's called BOX and it's there for your University if you want. http://www.internet2.edu/news/pr/2011.10.04.box.html

Wow, contempt for the non-university masses, much?

Anyway, I have two suggestions:

1) If you liked the way the Internet was (an ivory tower of academia), you should drop the current Internet, and get on Internet2.

2) There's more to the Internet than the World Wide Web. If you don't like the web, don't go there. Stay on Usenet and Gopher. Gopher has more servers than it used to, and you can access it via Lynx or Firefox addon. Or block non-.edu domains.

And please leave the no-pay, ad-supported Web to the rest of us.

Here you go.

Having seen your sig, I had to know - what is the bandwidth of a 747 filled with CD-ROMs?

So, I assumed a 747-8F, the latest freighter version. Payload is 148 tons, volume capacity is 21,347 cubic feet.
CDs weigh an average of 0.58 ounces, plus 0.11 ounces for a paper sleeve for a total weight of 0.69 ounces or 0.043 lbs.
A CD is 1.2 mm (0.047... inches) thick, the sleeve is about 2/250 or 0.008 inches thick, and the CD is 4.7 inches in diameter for a total volume of 0.96 cubic inches or 0.000555 cu. ft. I ignored the extra width and height of the sleeve.

So, in sum, ignoring other packaging issues, I figure the 747 can carry up to about 6,863,768 CDs by weight, or about 38,486,563 CDs by volume. Weight is thus the limiting factor. Using 750 MB CDs, that provides a packet size of 5,147,826,087 MB, at an average flight speed of 480 knots(552 mph, 889 km/hr). If we ignore loading and other ground time, and assume a typical flight from US east coast to US west coast of 3000 miles, that's about a 5.6 hour, or 20048 second flight. So we have a bandwidth of about 5147826087/20048 = 254234 MB/second or 1.94 terabits/second, within an order of magnitude given the sloppy estimating.

Not bad, but the latency really sucks. BTW - I note that the fastest internet II speeds are cited as 272,000 terabit-meters per second. So, using 5000000 meters (3000 miles), I figured that the 747 can do 9,698,256 terabit-meters per second, about 35 times faster.

Posting as AC, huh? Are you an NTRU Cryptosystems employee?

Here's a paper that surveys a number of quantum resistant cryptosystems. "NTRUEncrypt has also been found to be vulnerable to chosen ciphertext attacks based on decryption failuress [18, 21, 31, 38], but a padding scheme [30], which has provable security against these attacks, has been developed." "A comparatively greater number of problems have been found in NTRU-based signature schemes." "In 2006, it was shown by Nguyen that the unperturbed NTRUSign could be broken given only 400 signed messages [42]."

I'd say that the jury is still out...

Which ones? Googling for network speed records gives: http://www.internet2.edu/lsr/history.html and they're Linux...

http://www.internet2.edu

The problem with forking the internet is the question of what content is there going to be on it?

There currently is the Internet fork aka Internet 2 http://www.internet2.edu/network/ where people with content to share decided to create a new network and only allow people to play there that were of like mind. They won't be willing to let zinga or netflix come to I2 because those things are not inline with their goals and in the same way they are not going to make it available to Joe sixpack with a cable modem though admittedly there isn't much content on I2 that Joe would be interested in. But the cost to create a build out like I2 is huge and I don't see a private network or even a fork of the internet being able to build out to the point that the financial aspect would make sense. You can claim that once the people are there that the content will follow but really without the content the juicy center of the user base that advertisers like won't go there.

And if you think that getting things like video chat and video games to work when you are mixing clients is hard imagine if in the beginning you had to plug into your new ForkNet cable modem to be able to talk to your techy friends and into your Comcast modem to talk to your mom and dad who don't live close enough to a POP to get on the new internet fork.

I'm surprised no one has mentioned Shibboleth. Shib is pretty popular in the higher-ed space. There is a bit of a learning curve when first playing around with it, but once you get it up and going, it's very powerful. It does more than just your average SSO by providing federated authentication across organizations while maintaining user security. The project page is at shibboleth.internet2.edu

http://www.internet2.edu/ - yeah *that* worked...

Whatever happened to Internet2?

Here at UCLA (which participates in Internet2 and CENIC and some other organizations), it's not uncommon to see 40 Mbps download/upload in offices and 25 Mbps download/upload through the campus-wide WiFi for students. I can get WinXP SP3 in around 5 seconds...

Not to brag - I actually fear what might happen if some worm or hacker gets access to such high-speed network... :/

those tools seem pretty useful, but i don't know how user-friendly some of them are. personally, i'm looking for a tool to see if our ISP (at the office) is hijacking our DNS errors, or all of our computers are just infected with malware.

also, is anyone else seeing a bunch of "" characters on the Network Diagnostic Tester homepage? is my browser/system screwed up, or are there a bunch of a little boxes with "FF FD" in them scattered all over the page?

This is called eVLBI, and it is now done routinely to, e.g., determine the Earth's rotation (UT1).

From a networking standpoint, one interesting thing is that eVLBI requires high bandwidth (1 Gbps is typical), but can tolerate fairly high loss rates (because the actual cross correlation coefficients are rarely as high as 10^-3). This makes it an excellent candidate for an Internet scavenger service, where packets are sent at "less than best effort," i.e., with the understanding that they can be dropped if there is any congestion at all, so that eVLBI can use all available bandwidth without choking out other uses. The same technology may prove to be very useful for P2P services.

http://shibboleth.internet2.edu/

Yeah, I am aware of the I2 project. According to them IPv4, IPv6, and others are supported. Also, their list of 213 participating universities would make that about 5% of the total number of universities. It also begs the question of what subset of the participants are actually running IPv6?

Yeah, I am aware of the I2 project. According to them IPv4, IPv6, and others are supported. Also, their list of 213 participating universities would make that about 5% of the total number of universities. It also begs the question of what subset of the participants are actually running IPv6?

Um, yeah: http://www.internet2.edu/

http://www.internet2.edu/about/

TFA says Internet2 is "donating" bandwitdth to this project? Can we please focus if we're going to be serious about it? Can't BBN join Internet2 just like the 70 other corporations "leading the way to a new Internet"?

Oh sorry, these people all consider themselves CIO's and so this is the Second Management Unit coming up to speed. Excellent. Internet 2 should finally be complete in one more Mythical Man Month. Just wait!

-Matt

P.S. Where did these guys rematerialize from? They got eaten and re-eaten back in the 90's! BBN has been shat out from the telecom industry like a seed from a bird. (And that bird seems to have diarrhea.)

This should help fuel your paranoid delusions.

...which already exists.

FreeBSD changes its APIs in every major version, so you're supporting my argument by naming it as a fast operating system. Even then, the developers regularly admit that the USB stack still sucks, for instance. Not enough gets fixed.

What do you mean, which benchmarks does Linux top? Which ones doesn't it top? SMP scalability has been leading since 2.6, and the closest free software contender is FreeBSD 7, released years later. Most of the leading Internet2 LSR records are Linux, maxing out network hardware. http://www.internet2.edu/lsr/history.html

Linux DOMINATES the high-end SMP space. Windows barely makes an appearance at all. Windows doesn't even /boot/ on advanced processors like the Cell. BSD does, but has no support for any of the coprocessors.

"Other OSes" don't achieve these feats, and those that come close have this API flexibility requirement as well. All of the evidence supports my argument, and the Linux developers at large agree.

Mod parent insightful, instead of funny. Various people and some papers have suggested that upgrading network capacity is a better way to handle high traffic than trying to mess with QoS, because 1. it's cheaper 2. it actually works, which isn't really proven to be the case for QoS on a large ISP level network.

I think it is a good time to reply here and link to the paper that prompted me to write the OP in this thread. I think there was even a slashdot story about it.

It's far from the only TCP measurement tool. Besides, don't you think attacking ICANN's corporate land-grab from within is a better use of someone's time than spinning version numbers on a stable program that talks to a stable API?

http://www.internet2.edu/pubs/networkmap.pdf Good thing they made a little turning in the detroit - cleveland connection otherwise Canada would be on the internet2.

Looking through the list I wonder how many of these patents have been contested and/or how many of these patent applications should be rejected. For instance, under the "Web Browser Federated Sign-On Protocol" Microsoft lists patent applications 2005-0223217-A1 and 2006-0112422-A1. There exists prior art for 2005-0223217-A1 claim 14 and further in A-Select and I think for the rest of the claims in Shibboleth. I would have to study 2006-0112422-A1 more in detail, but it looks like about half of the claims there also have prior art in those systems.

http://www.internet2.edu/

High Performance Bulk Data Transfers Using the SSH Protocol

These basic ideas are already being worked on with such systems as myVocs (pdf), IAMSuite, and CoManage. It is an idea whose time has come due. It's basically about the web maturing and adopting system boundaries (however loose or tightly you want to define them). It's a similar transition from DOS->Win->NT (or any batch to multitask migration you want to draw a parallel to). The web is about like DOS right now.

Doesn't this already exist? I mean, seriously, how many parallel projects do we need to do the same thing?

http://www.internet2.edu/network/library/deploymen t_phases.pdf

http://www.internet2.edu/network/deployment.html

http://www.internet2.edu/network/library/deploymen t_phases.pdf

http://www.internet2.edu/network/deployment.html

A map of Internet2 without MIT? Somethings not right here...

This list [PDF] is better, and while it confirms that this particular college isn't on Internet2, University of Maine is a member. Your sleepy home state isn't entirely left out of the fun it seems.

http://en.wikipedia.org/wiki/Internet2

There have been times when the media have reported on a network called "Internet2." This is misleading since Internet2 is in fact a consortium and not a computer network. "Internet2" is sometimes used, albeit a misnomer, for the Abilene Network.

So I take a look at Abilenes website and find this map: http://abilene.internet2.edu/peernetworks/domestic .html

From what I can see here, It does not look like the Internet2 network reaches as far north as Maine.

Having been born and raised in Maine (living in CA now), this really comes as no surprise to me. There are only about 1.2 million people there, certainly considered the "last mile" by most providers (cell, cable, dsl...)

The University of Kansas will capitulate to anyone that threatens them. The chancellor is inept and doesn't care about the students. The IT Security Office is manned by a bunch of lunatics (all the awesome old guys from Academic Computing Services are gone) that will do anything in the name of "security." Also at this school, the attitude towards students is that they are always wrong, and the Great and Holy Chancellor and his Apostles in the administration are always right. They will destroy students to appease the RIAA, mark my words.

I'm intrigued that the RIAA would be threatening Internet2 members. Notice that the RIAA is a member of Internet2. What possible positive contribution can they make? Look at the rest of the members. I think that if I were a member institution threatened by them, I would petition for their expulsion from the network.

I have looked at the partner sites for the project http://international.internet2.edu/partners/ there doesn't seem to be much progress outside of the US. Are there plans for any that anybody is aware of?

The press release: https://mail.internet2.edu/wws/arc/i2-news/2007-06 /msg00003.html

See http://www.internet2.edu/network/library/deploymen t_phases.pdf

Almost as interested as what is covered is what is not yet covered. Does anyone know how long they expect this to take to fully implement?

It won't be cheap, though.

The entire Internet2 network has very limited or no alternate routing capability.

Indeed, http://www.internet2.edu/pubs/200402-POS-AN.pdf seems to confirm that only the top-tier has any redundancy at all, and what it does have is quite limited. Most likely your explanation of it being cost-related is spot on.

Maybe the network topology info at http://noc.net.internet2.edu/i2network/maps.html will provide more info.

Indeed, http://www.internet2.edu/pubs/200402-POS-AN.pdf seems to confirm that only the top-tier has any redundancy at all, and what it does have is quite limited. Most likely your explanation of it being cost-related is spot on.

Maybe the network topology info at http://noc.net.internet2.edu/i2network/maps.html will provide more info.

For quick testing customer connections I setup an NDT server. http://e2epi.internet2.edu/ndt/
It works great in Firefox, but requires a real java installation. The last time I checked, it didn't work at all in IE. My laptop with an integrated Gig-E port gets about 800megs in and 200megs out running XP and 800megs in and 600 megs out running linux. On a shared Gig-E connection here in the office it's usually within a meg or two on repeated tests. I figure that's close enough for testing 10-100meg customer connections. At customer sites it can usually give me the link CIR to within a meg or two. And it'll point out probable duplex mis-matches too.

For more extensive testing we have Ixia Chariot and a pair of their chassis. http://www.ixiacom.com/
It's expensive and has lots of options. The Voip and QoS modules are coming in very handy right now, but we've had a lot of trouble with support since Ixia took over from NetIQ. The hardware is impressive but they've got Windows XP embedded on it to run the embedded linux systems that run the ports. The software endpoints are available for a variety of OS's. The pricing is based on the number of concurrent pairs you want to use for testing. I'm still not happy about how much we spent (and still spend on maintenance) for this stuff, but my PHB likes the pretty graphs so I guess that's all that really matters. We mostly use it for multi-hour validation testing when new sites and links get installed and sometimes when we need to verify that a vendor has in fact upgraded the bandwidth on a circuit.

The End to End Performance Initiative has a knoppix live CD image you can download that includes test tools that may help. I'm in the process of deploying these tools around my network now.

I've not tried to push a full gig with them (yet), but they seem to work better than anything else I've found so far...

http://e2epi.internet2.edu/network-performance-too lkit.html is the URL.