5% of the Net is Unreachable

dasheiff writes "A BBC Story says US researchers reveal that up to 5% of the internet is completely unreachable. However the most interesting part is that they reported that many of the lost net sites flare into life briefly when being used to send spam or to launch attacks on other parts of the net."

Comments

Unreachable?

[WinstonSmith]

If it is unreachable, is it really part of the Internet?

When I turn off my router, I don't really consider my home machines part of the Internet even though they are running and connected by a physicall wire.

Re:Unreachable?

[d5w]

If it is unreachable, is it really part of the Internet?

Check out the Arbor Networks presentation the BBC is referring to. Their definition of "dark address spaces" is

"The range of topology accessible from one provider, but unreachable via one or more competitor networks"

So, yes, these addresses are reachable by someone, just not by everyone.

Re:Unreachable?

[Kirruth]

You can reach these "dark spaces" if you know what you are doing. The simplest way is to use an http proxy (or tracert host) in another part of the net, or just use another isp. They are not unreachable in that sense, even though the default route from where you are may not work.

Spammers or system crackers often seem to do the trick of hacking into a set of home user broadband machines, I guess using a trojan or worm, turning them into a chain of proxies, then nailing the router between the last of the proxies and the rest of the net. In this way they make their own dark space.

Sites behind NAT

[category9]

I'd wager a great deal more than 5% is inaccessible if you count all the home sites locked away behind nat firewalls. Once we all start getting hundreds of IPv6 addresses at home, we'll start to see hundreds more small home/user sites popping up. This could greatly change the structure of the net, once again breaking away from the central information resources we are beggining to solely rely on and start using small independent resources much more.

Re:Sites behind NAT

[madcoder47]

IPv6 is going to take forever to implement! The whole net infrastructure is IPv4, and to upgrade software, NICs, Routers, Firmware-of-net-devices and the like will cost fortunes and be a lengthy switch. sure I'd love to see the day when we all can have 100s of IPs, but i dont think it will happen soon.

Ummm....

[NiftyNews]

How exactly does one define any part of the net as "unreachable?" Doesn't the term "internet" imply that it is available on the network?

Seems kinda silly if you ask me. Why not declare that 59.28% of the internet is unreachable? Why not 600%? They're all equally unprovable and meaningless ;)

People get paid to run "ping"? This is research?

[Wakko Warner]

Seriously, I could've told you that 5% of the net is unreachable at any given time. It's called "PPP Connections". This is some sort of breakthrough research?

- A.P.

Re:This leads to an interesting possibility

[benb]

> Seems to me that you could make some progress
> against the spam by simply refusing any email
> from a domain that hadn't been recognized on
> the net for at least several days or maybe weeks.

They will just add sites (unused) 2 weeks before the spam-attack, but you will hurt honest users and admins a lot, because you just tremendiously increased the time it takes to move/add sites.

You look to me like those "copy protection" guys. You are participating in a cat-and-mouse game, but don't care about hurting other people's interests for your cause, without even achieving it.

Spam protection must never hurt honest users.

Re:Content-free article

[satch89450]

What worries me is that it took someone three years to figure this out...

I think you may have jumped to a wrong conclusion here. It didn't take three years to figure out that spammers play around with unsecured routers. It took three years to prove via experiment and measurement the extent of the problem, and to quantify the extent of the problem.

When the little boy has cried "Wolf!" often enough, the lone cry is quickly ignored. When the little boy then yells "Wolf, range 600, bearing 219" the cry takes on a bit more significance, don't you think?

If you can't measure it, it's opinion not science. (No, I can't find who said it first -- it's not original with me.)

Re:Content-free article

[Anonymous Coward]

If you can measure it, it's not science, it's established fact.

Science is the process by which hypothesis are transformed into established facts. Those established facts are used by scientists but they are not science. Science is that leading edge vanguard that is at the tip of knowledge asking 'why.'

In other words, the people teaching school children Newtonian Physics are not scientists. They are educators.

In the final analysis who cares what some turgid bromide-spewer like Heinlein has to say?

Re:Spammers, may they rest in the damnation of hel

[geekoid]

The fact the the company works for, may or may not making tools used for spamming, the outcry from the community is over something for more important the anti-spamming, basic civil rights.

I's rather get 50 pieces of spam in my email, then 1 piece of junk snail mail.

The way to slowdown spamming is control, not outlaw. If you outlaw spamming, you will be outlawing anything similiar to it regardless of intentions. This will have an impact on free speech, on others beside spammers.

We also need an official definition of "spam" put before congress, before ANY laws or actions are taken.
Is it spam if it's primary use is to make money?
is sending you a joke spam?
Is sending you a political announcement spam?
IS sending you any email you didn't explicitly ask for spam?

Don't get me wrong, I don't like unsolicited email from certain groups, I just feel we need to exam and define what spam is, and consider possible unexpected consquence before we make laws.

First Saturday of Every Quarter

[Multics]

How about everyone spend an hour or two on the first Saturday of every quarter working on 'hurting' all the SPAMMERS in their mailbox?

Hurt could be legal (complaints, blocking, etc), quasi-legal (nmap, ping attacks, etc) or illegal (kill the bastards and drag their guts down the block as an example of what could happen to spammers in the future). Let your rules of engagement be your guide.

If we all spent 1-2 hours on this four times in 2002, I'll guess that there would be fewer spammers in the trade by the end of the year, not more.

Thoughts? I'll stay legal for the moment.

-- Multics

Re:ummmm

[billn]

Yes, this is a reprint of an older story, found here.

Laws are the wrong answer

[HiThere]

Laws are definitely the wrong answer.

The right answer is a configurable e-mail filtering system. With certain pre-programmed options. And easy customization. And PERHAPS a neural net that can learn what it considered spam (or, perhaps better?, not spam).

It needs to be cross-platform. It needs to be able to work with MSIE. It may be MSIEvil, but it's the predominating e-mail recipient.

This doesn't get around the need to receive the verfluct stuff, but if the job is done well enough, it will get around the yammering for more laws. People should be able to set their own priorities. (If it were easy enough, I'd automatically reject anything that was predominately non-indoeuropean letters. I don't read Japanese, Chinese, and whatever those other languages are, so it would be nice to avoid them. But sofar I haven't bothered figuring out how to reject them before reading.)

I can't even imagine any way to reject the garbage without receiving it, except rejecting based on ISP, sender, addressee (e.g., list suppressed), subject, or date. And that's not usually enough to go on. But sometimes it is, and it would be nice to delete those before downloading them.
.