Slashdot Mirror

Well, yes, it's technically voluntary for ISPs to implement the blacklist. But the previous government essentially threatened to make it mandatory unless every ISP 'voluntarily' adopted it.

In the video as linked to in my other post, note in particular around 32 min, when this guy - whose business provides the best native consumer IPv6 connectivity in the UK - points out how awkward Google were with him.

Good lord, you've *got* to be kidding. Google has done more to push v6 than virtually any other content provider out there.

What has Google done to push IPv6, i.e. in what way has it demonstrated feature parity with IPv4 or benefits over IPv4? The fact that Google is considered at the forefront is a sign of how little /anyone/ has done in the public space - and it hasn't done more than any number of content providers, smaller ISPs (that's the way to do it!), etc which already are involved in IPv6.

But, to answer your question, why not start with the sixxs coolstuff? Note in particular that multicast is demonstrated, and that promotional IPv6 services are offered (newsgroups) even when the service could be provided as easily with IPv4. It's a very minor marketing effort, but it's still got substance to it much greater in proportion to size than Google has (hitherto) demonstrated.

What is this "choice" you're referring to? Because last I checked, users would have a choice: run dual-stack or don't.

The choice to use IPv4 or IPv6 (with extra features, where IPv6 can be so exploited) sites.

So now you're saying that, to fix this problem, we should expect those consumers to go upgrade their routers? It's just not gonna happen (outside the normal obsolescence cycle).

And how exactly are ISPs going to get Approved[tm] by Google if most of their customers have broken routers which ruin the fun once AAAA records are sent?

while simultaneously transparently enabling v6 for those users who have an ISP that's on the ball

But this will just mean a decrease in performance until *end-to-end* routing/bandwidth for IPv6 is as good as for IPv4. You might not get the n second timeout - the nature of the lameness just changes. An ISP doesn't either have good or bad IPv6 connectivity to the whole world!

unless content providers start populating the v6 web, it'll continue to go nowhere fast,

Unless they start populating it with something users want. Even if it's just feature parity combined with the vanity stage I mentioned a few posts up, where you're visiting either www.blah.com or www.legacy-ipv4.blah.com. But making it up to Google rather than the customers whether services in general are good enough or better on IPv6 is daft.

https://www.linx.net/pubtools/trafficstats.html?stats=day
(The London Internet Exchange (LINK) traffic graph, the second largest peering point in the world, so it should be a good average.)

It's shifting 230Gbits/s at peak, and less than half that amount from midnight to about 8am -- that's probably why the UK ISPs that change per GB don't limit bandwith use at night -- it's spare capacity.

How can refusal to cooperate with an investigation itself be illegal?

Because the law makers chose to make it so. The rights to silence, non-self-incrimination etc. have all been curtailed, usually under the guise of anti-terrorism but also driving laws.

It is, for example, an offence (in the UK) to refuse to tell the police who was driving your car at a particular time (for speed camera purposes).

Some links:

RIPA activation: http://publicaffairs.linx.net/news/?p=513

Driving: http://news.bbc.co.uk/1/hi/uk/5383726.stm

General rights to silence: http://www.yourrights.org.uk/your-rights/chapters/ the-rights-of-suspects/the-rights-of-suspects-in-t he-police-station/curtailment-of-the-right-to-sile nce.shtml

Offence to fail to cooperate with terrorism stop-and-search (note: no suspicion is required for this type of stop and search anymore): http://www.yourrights.org.uk/your-rights/chapters/ the-rights-of-suspects/anti-terrorism-powers/anti- terrorism-powers.shtml

For an example of this visit https://www.linx.net/www_public/our_members/peerin g_matrix/ that shows the peering matrix at the London Internet Exchange.

AMS-IX has more members than LINX (list of LINX members http://www.linx.net/members/index.thtml compared to AMS-IX http://www.ams-ix.net/connected/ but that's probably due to the fact that to become a LINX member you have to fork out the best part of £10,000 (GBP), whereas I'm pretty sure (although feel free to correct me) AMS-IX is no where near that amount.

Most ISPs value their LINX peering much higher than any other peering (how many web hosts have you seen state they are a member of AMS-IX before they state LINX? Also LINX is generally seen by most ISPs are the main peering point between US and Europe (not even just the UK). Large ISPs in the UK (non-Telco ones) such as Claranet, Demon, Nildram etc have 'fatter' pipes to LINX rather than other European peering points (such as PARIX or DECIX).

The AMS-IX is the largest Internet Exchange / NAP in Europe

It is? I was always under the impression LINX held that honour..

You can have one person controlling the power for a building.
About 7 years ago, a company I used to work for had one of their major datacenters in a room in Telehouse Docklands (in London, in the UK). It's a big building and it also hosts the London Internet Neutral eXchange. One day half the building lost power. An electrical technician in the basement turned the wrong switch and then fainted when he realised what he had done. This meant it took a while to turn the power back on, because you don't just walk up and touch an unconscious person next to an anomalous electrical power situation.
We had in-room UPS but they had to be slaved to the main power switch so we couldn't have an independent electrical fire in the room, so they went off when the power was switched off.
Most of our servers recovered OK, a couple did not.

As for it taking some time to get the servers back online, databases always suck for that. None of them seem to start up reliably all the time without DBA intervention; not PostgreSQL, not Oracle, not Sybase. They all have a tendency to eat their innards every so often after an unplanned shutdown.

The LINX Best Current Practice on Unsolicited Bulk E-mail ("the spam BCP") is carefully written so as to avoid being a way to create denial of service attacks.

LINX does not adjudicate complaints; our ISPs members do. You can complain to an ISP for tolerating spamvertised web sites just like you can complain to them for tolerating someone sending spam. If they follow Best Practice they will cut off the web site if, only if, and not before they satisfy themselves that the spam was sent by or with the consent of the web site owner.

Of course, it is possible that they could get it wrong; miscarriages of justice do occur in every area of life. This is not a reason not to have any rules at all. It is up to the ISP to take care when considering a complaint so as not to cut their customers off without good reason. Naturally, some will consider this an unnecessary delay - and even evidence that the ISP is not serious about cancelling the account. Well, it's not possible to please everybody all the time; you've just got to craft the best policy you can and run with it.

Malcolm Hutty
LINX Regulation Officer.

The LINX Best Current Practice on Unsolicited Bulk E-mail ("the spam BCP") is carefully written so as to avoid being a way to create denial of service attacks.

LINX does not adjudicate complaints; our ISPs members do. You can complain to an ISP for tolerating spamvertised web sites just like you can complain to them for tolerating someone sending spam. If they follow Best Practice they will cut off the web site if, only if, and not before they satisfy themselves that the spam was sent by or with the consent of the web site owner.

Of course, it is possible that they could get it wrong; miscarriages of justice do occur in every area of life. This is not a reason not to have any rules at all. It is up to the ISP to take care when considering a complaint so as not to cut their customers off without good reason. Naturally, some will consider this an unnecessary delay - and even evidence that the ISP is not serious about cancelling the account. Well, it's not possible to please everybody all the time; you've just got to craft the best policy you can and run with it.

Malcolm Hutty
LINX Regulation Officer.

For another the only people who will noticed the difference between mirror.ac.uk and any european FTP site are those on janet (joint academic network).

It's hard to believe there is 'heavy' use of IPv6 when the dedicated IPv6 exchange in the UK peaks at 4Mbit/s of traffic and the LINX exchange in London has >30Gbit/s of IPv4 traffic

https://lg.ipv6.btexact.com/lgmrtg/hopper-day.html

http://www.linx.net/tools/stats/index.thtml

I've been looking at the traffic statistics of some european internet exchanges and none of them seem to have more traffic then usual.

The 2 reasons I can think of:
1) There is no extra traffic. (maybe a little)
2) All traffic goes thru transit providers (cause it should go to U.S.)

If there are over 200.000 machines infected by MyDoom there should be noticable extra traffic.

References
Amsterdam
London
Brussel
Paris

On the topic of economics, here's some numbers I just whipped up:

LINX, the London Internet Exchange, deals with around 15Gb/s on average. 60% of that is a little over 1GB/s. Let's be conservative and call it 0.9GB/s. That's 77,000 GB/day (rounding down again), which is all P2P (so they claim).

That amount of data comes to 7,000,000 10MB music tracks, or 77,000 1GB video files. Now, Apple have proven that Joe Public will pay a reasonable amount for media offered legitimately, so let's assume that 10% of this P2P traffic could be converted to legitimate sales of music. So, assuming £0.60 (=$1) for a music track, or £4 (=$6.50) for a movie, that's a potential:

£42,000/day -> £15million/year of music sales ($25m/year) or £31,000/day -> £11million/year of movie sales ($18m/year)

And that's just in the UK!

Now, I know this isn't the point of the article, but there's a huge potential for the media companies we all moan about so much to make money, *and* for us to get the content we want (sans 9 tracks of filler for one good track) at the same time, and they're doing almost nothing about it. What am I missing?

(OK, so I guess someone who knows more about bandwidth usage and economics than me can make these figures more realistic, but you get an order of magnitude...)

PhilRod

Have a look at the traffic statistics of the public peering points in Europe:

LINX - London - 25Gbit/s
AMSIX - Amsterdam - 11Gbit/s
DECIX - Frankfurt - 10Gbit/s

If you look at it most of them double traffic even faster than in 12 month. I think it's closer to 9 month.

--
Andre

Australia really needs some not-for-profit (government sponsored?) peering points (IX's) like they do here in the UK (Linx, Lipex). Having large ISP's in control of the peering game will inevitably make the rules unfair.

Of course, even if someone could knock all the root servers over, the net as we know it wouldn't stop working instantly. That's what the time to live value is for :)

Janet's connections are not, and were not, solely KPNQwest. There's LINX, GÉANT, and Teleglobe as well.

If you take a look at the LINX peering matrix (warning: insanely large HTML table), you'll see that JAnet (JNT) a.k.a. UKERNA is peered with many other providers, including UUNet and PSI, so no, I don't think they're going to disappear overnight.

Clarification:

ja.net charges for transatlantic bandwidth partly because a great deal of ja.net's costs were on paying for transatlantic lines. Anything that doesn't travel via these lines (ie via LINX or GEANT is free which means most resources in the UK or europe.

Ja.net has also mitigated the need to use so much transatlantic traffic through the use of mirror.ac.uk and the National ja.net webcache.
If only more people were to use these, a number of smaller UK academic institutions are not aware of a number of services that ja.net can provide for them but this is changing in recent years.

No it doesn't, at least not in the UK, and I'm pretty confident US bandwidth is cheaper than here. How much bandwidth do you expect to need? For a couple of thousand pounds a month, you'll get a 100Mb/s pipe direct to a LINX backbone provider in Telehouse. For a few thousand more, you get gigabit (yes, true gigabit internet access). I couldn't believe how cheap bandwidth had become when we were looking at it over the past few months.

Yes, but that's not what they're doing. If you look at the matrix of UK peering agreements, you'll notice that UUNet doesn't peer with quite a lot of the other Linx members. All Linx members are backbone ISPs, not small local outfits, but UUNet (and some of the other larger players) still refuse to peer. And that's refuse to peer, not "we'll peer with you if you pay us". Because at the end of the day, refusing to peer with the smaller guys protects their monopoly, and that's far more valuable to them than the income they'd get by selling peering agreements.