Password Gropers Hit Peak Stupid, Take the Spamtrap Bait

badger.foo (447981) writes Peter Hansteen reports that a new distributed and slow-moving password guessing effort is underway, much like the earlier reports, but this time with a twist: The users they are trying to access do not exist. Instead, they're taken from the bsdly.net spamtrap address list, where all listed email addresses are guaranteed to be invalid in their listed domains. There is a tiny chance that this is an elaborate prank or joke, but it's more likely that via excessive automation, the password gropers have finally hit Peak Stupid.

Comments

This guy might be overvaluing his files

[damn_registrars]

I expect his file was probably indexed by a search engine (he does talk about it fairly often in his blog) and the botnet found it there. The botnet isn't smart enough to know that the email addresses aren't real - it only knows they are valid - so it went ahead and went for it. Hell if you were looking to compromise email addresses for your own nefarious purposes and had a small army of compromised PCs to attempt the password hacking, you wouldn't care if you were attempting to access valid addresses or not.

Re:This guy might be overvaluing his files

[s13g3]

I designed a honeypot built on similar principles at the last data center I worked for, whereby I had at least two different VM's comprising at least two different OS' on each and every subnet on our network.

Using a custom implementation of PSAD and a bunch of PERL, the basic idea was that any time a specific IP (external *or* internal) scanned more than eight ports per IP across two or more subnets, it was unquestionably an illegitimate scan of our network, and the IP originating the scan in question was immediately submitted for null routing, because nobody could possibly have a legitimate reason for doing such a scan.

Port scans from internal IP's, along with those matching other patterns (such as multiple scans within a single subnet or attempting certain exploits/attacks that can be deduced from snort's output in /var/log/messages, like the slammer worm, etc.) were output to a file that was reviewed daily, and could then be fed either in whole or in part(s) to a script that would process the desired actions. Before I knew it, I was blackholing hundreds or even thousands of addresses a day... ~70% of which were from China Telecom, followed immediately by Russia, Brazil, and Moldova, with less than 5% of attacks originating from U.S. or European addresses. The number of compromised customer servers on our network plummeted, along with a corresponding and by-no-means-insignificant dip in network traffic.

What got me started on this project was that, among other things, hackers were scanning our network for Plesk's default admin login port (as Plesk at that time *had* a default admin login and password), and any time they got a response from port 8443 on an IP that previously did not have that port open, they would jump in and root new installs often before the customer ever logged in for the first time. Needless to say, I put an end to that nonsense.

However, calling spammers dumb as others have above is probably a mistake: they can often be fairly smart, but what they really are - usually - is Peak Lazy, and are aiming for low hanging fruit. Eventually, the more sophisticated ones will create or adapt new techniques to defeat - or at least cope with - this particular methodology, and the cat-and-mouse-arms-race game of security will continue on as it always has, with one side or the other evolving new defenses or offenses, and the other evolving an appropriate response. The fact that a particular batch of spammers got caught and will find the emails from their current spam campaigns not reaching their intended audience on this go round will only slow them down for a time on the domains this list covers, but to say the spammers have hit "Peak Stupid" as a result of excessive automation is, in fact, an NP-Dumb analysis.

One script kiddie made a mistake

[Nimey]

so now they've all hit peak stupid.

I'm not sure it's the script kiddies that have hit that or the submitter and editor.

Re:One script kiddie made a mistake

[Noah Haders]

unfortunately, it's unlikely to be "peak stupid." This would imply that stupidity has hit a maximum and things are only going to get less and less stupid as we move forward. Never undervalue humanity's capability to get more and more stupid as time goes on.

although to be fair, you could call the nuclear arms race "peak stupid" because humanity was flirting with destroying all human existence. n00b spammers have no chance of being this stupid, and hopefully we will never be so stupid again.

Don't be silly

[Kierthos]

There's no such thing as 'Peak Stupid'. Every time someone gets to the top of the current peak, the fog clears and another mountain of stupid looms in front of them.

Re:Don't be silly

[alex67500]

A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools.
Douglas Adams

Re:Don't be silly

[Thanshin]

There's no such thing as 'Peak Stupid'. Every time someone gets to the top of the current peak, the fog clears and another mountain of stupid looms in front of them.

A phenomenon well documented in the study "the unpeakability of stupid".

Peak Stupid

[wasteoid]

So is trying so hard to coin a phrase like "peak stupid".

Re:Editors

[Hsien-Ko]

These moron editer's should better there English.

Well

[Spad]

While reading this story I accidentally peak stupid.

Re:Well

[TapeCutter]

Obviously now is the time to sell stupid.

Isn't "Peak Stupid" writing about it.

[Chrisq]

The script kiddies are wasting time and resources looking for non existent email addresses. Wouldn't it be better to let them get on with it rather than tell them exactly where a whole list of email addresses that they needn't check can be downloaded?

Maybe this can be used against the bots

[ZorinLynx]

Populate the net with files like this full of E-mail addresses that are not valid. Have dummy accounts on the appropriate servers that will accept the logins, allow the spambots to think they're successfully sending E-mails when in fact they're all going into the bit bucket.

For added effect, make the servers respond v e r y s l o w l y under these accounts, taking tens of seconds to "send" the E-mail, a minute or so to log in, etc. Basically, slow the spam bots down and waste their time. Of course, the bots will probably eventually evolve to detect such shenanigans, but why make spammers' jobs easy? :)

How fucking stupid are you

[Mr 44]

This is great news for stopping this particular batch of spam.

You just posted the same point twice in this thread, and its completely wrong both times, and shows a total lack of reading comprehension on your part.

They are NOT emailing these addresses, they are attempting to log in to them.

Read the fucking summary, at least. You are what's wrong with the internet.

It's been done. (teergrube)

[oneiros27]

There's even a term for this, teergrube.

An ISP that I worked for in the 1990s used to do this (dcr.net, owned by Drew Curtis, of fark.com fame).

We had some code that would look for blatant e-mail harvesters, and would SLOWLY return random bogus e-mail addresses ... wait a couple seconds, spit out an address ... etc. The page at the top even had warnings that the page was completely bogus.

At first, all of the e-mail addresses were all in our domain (but not our real mail server), but I went and added some code that would look up the connecting IP's network (I think I used whois.ra.net), and would also include '{abuse,postmaster}@(network)' and again for the network's upstream providers.

I can't remember if the bogus mail server was also the box that we had set up so that if *anything* tried touching it, it'd blackhole the connecting IP at our external router, if it was a teergrube itself.