|
Grex > Coop12 > #165: What to do about spam forged with @cyberspace.org return path? | |
|
| Author |
Message |
mdw
|
|
What to do about spam forged with @cyberspace.org return path?
|
Jan 11 08:49 UTC 2003 |
A small group of spammers has come up with an annoying new variation on
the spam they send out. Here's the apparent situation:
#1 one (or a small group) of spammers have bought a CD and
software to send spam.
#2 the spammer composes his spam in HTML.
#3 the software goes out, probably uses a dialup IP address to find
some sort of unwitting proxy server or open relay, and sends
out the spam. It sends it out to a large list of people,
sorted by domain, probably one message per domain, and then
using a fake forged return path that points to some random name @
cyberspace.org It is efficient to send one message to a domain,
so the spammer can get away with a low-speed dialup connection.
#4 the open relay or proxy server, which very likely failed to record
the dialup IP address, then sends the spam onwards to the
target domain.
#5 the target domain attempts to deliver the spam. The delivery fails
for some users, either because their mailbox does not exist,
the mailbox is full, or anti-spam rules stop the HTML spam.
#6 the target domain then sends at least one bounce message to the
return path, which points to us. Each bounce contains the
entire original message plus lots of information on the
bounce. This increases the size of the bounce by a factor of
2. Some domains instead deliver several hundred bounce
messages. This further multiplies the effect by one or two
orders of magnitude. There are at least 2 of these spammers,
another factor of 2. So we see at least 400 times as much
traffic over our shared DSL connection as the spammer generated
over his unshared dialup connection. The spammers have
been at this a number of days. So far, this resulted in filling
up at least 2 mailboxes with 1 M worth of bounces, and probably
a lot more text has failed to be delivered.
Some options I ccan think of:
#1 ignore the problem.
It might go away, or it might increase in magnitude in time.
If "@cyberspace.org" addresses become common on spam:
a. any address used by a spammer becomes useless
to its legitimate owner on grex, because
his mailbox instantly floods with spam bounces.
b. our network connection will become useless
a. other sites might just start blocking all incoming
mail addresses as being from "@cyberspace.org"
regardless of legitimacy.
#2 block bounces for spam not originated on grex.
This only helps with 1a.
#3 change mailbox syntax or domain name.
Changing mailbox syntax breaks all existing names. Possible syntax:
remmers+@cyberspace.org works
remmers@cyberspace.org does not work.
major user education headache.
Keeping the same domain means we still see bounces & have site
block issues.
Changing the domain means giving up "@cyberspace.org". How
attached are people to this?
We might be able to keep:
"@grex.cyberspace.org" which might be of some attraction.
Or we could add "@mail.cyberspace.org" which might argueably
be the right thing to do.
#4 complain to the proxy site, ISP owning the web site, etc.
Everybody does that with spam already. The spammer is probably
fully prepared to deal with the probable consequences of this.
I'm not at all convinced this is even worth attempting.
It may even make it harder to proceed with legal or criminal
complaints - by causing the spammer to flee scott-free, and
giving him a chance to multiply.
#5 seek legal action against the spammer proper.
Seeking legal action gives us rights we would not ordinarily
have. For instance, in a law suit, I *think* we can subpoena
or otherwise cause the ISP to cough up the name or other
contact information of the spammer. Legally, the ISP cannot
otherwise just give us that information; there are various
privacy laws that result in this. Cons: hiring a lawyer can be
expensive. Some work can be cheap or free. An initial
consultation is free. A 'cease and desist' letter should only
be $100 or so. Actual court costs are relatively low.
Actuallying going to court can cost $thousands. If we go that
route, we'll either need major fund raisers, or we'll need to
find lawyers who can work for much less cost or pro bono. This
is not necessarily unreasonable; as a 501c3 corporation that is
"well-known", there are even organizations (legal aid
societies) that actually specialize in providing free or
low-cost legal coverage for such cases, so we have decent odds
of finding such a person, given sufficient begging. In a civil
suit, we merely have to find "compelling" evidence the spammer
did it - we don't have to prove anything "beyond a shadow of a
doubt". If we win a civil suit, there's some chance of
recovering at least part of any money we invest in such a
suit.
#6 file a criminal complaint.
Michigan has a tough "fradulent use of computer resources" law.
pros: somebody else does all the work. Free. Tough penalties.
cons: not clear the prosecutor will care. They may not have
the experience or interest to successfully prosecute such a
case. Also, stricter standards for evidence make conviction
harder.
I don't want people to get too tied up in the technical aspects here.
I'll enter a different item in garage in case people want to talk about
that. Here, I want people people to think in terms of the policy issues.
ie,
do we care about internet mail accessibility?
do we value "@cyberspace.org" as our mail domain name?
do we want to bother trying the usual spam complaint?
are we willing to pursue legal remedies for people fradulently
infringing on "@cyberspace.org"?
|
| 26 responses total. |
glenda
|
|
response 1 of 26:
|
Jan 11 10:36 UTC 2003 |
Re #3: some of the bounce messages I have received showed
"glenda@grex.cypberspace.org" as the address. Sorry that I didn't keep them,
I sort of automatically delete spam. I did a quick look at these since they
said returned message on them to look at the headers. If I get any more I
will save them and pass them on.
|
mdw
|
|
response 2 of 26:
|
Jan 11 11:20 UTC 2003 |
If the "return path" is set to glenda than I am very interested.
If it's the To: header I'm not at all interested, and if it's the From:
header I am only mildly interested -- send it to uce.
|
gelinas
|
|
response 3 of 26:
|
Jan 11 16:04 UTC 2003 |
Certain virus software propagates itself by e-mail, as many of us know.
Originally, it would use the e-mail address of the owner/user of the
originating machine as the "From:" for its messages. Later on, it started
choosing addresses out of the addressbook on the machine and using one
or more of those address as the "From:" for its messages.
From Marcus' descriptions so far, this does not seem to be virus-related.
We may need to prove this, though. The easiest way to do so is to answer
negatively to "Did the address getting the rejection notice ever exist?"
Can we answer that question, at all?
For the rest of Marcus' questions, I want to hear what others think before
offering my own opinions.
|
carson
|
|
response 4 of 26:
|
Jan 11 17:22 UTC 2003 |
(another con to the legal action scenario is that Grex could find it
difficult to pursue action against an offshore spammer. that said, I
don't think we're going to find a perfect, one-size-fits-all solution.
however, we should still do *something*.)
|
other
|
|
response 5 of 26:
|
Jan 11 19:06 UTC 2003 |
As I mentioned in another item, I have contacted the Michigan Attorney
General's office to ask for information about filing a complaint. The
next response from me will contain the content of that exchange. I'll
hide it so those not interested won't have to scroll through multiple
screens.
|
other
|
|
response 6 of 26:
|
Jan 11 19:14 UTC 2003 |
View hidden response.
|
gelinas
|
|
response 7 of 26:
|
Jan 11 19:30 UTC 2003 |
(As a reminder, in bbs, at the "Respond, pass, forget, quit, or ? for more
options?" prompt, enter "set noforget" and then "only 6" to see Eric's
response. You will probably then want to "set forget", so you don't have to
see all of the other things you've forgotten.)
|
gull
|
|
response 8 of 26:
|
Jan 12 03:41 UTC 2003 |
Are they even doing anything illegal? Last I heard there were no laws
against spamming. It may be hard to win a case like this. (That may not
matter, though -- a cease-and-desist letter can be pretty effective
regardless of whether or not it has a legal leg to stand on.)
|
mdw
|
|
response 9 of 26:
|
Jan 12 03:54 UTC 2003 |
This is not a virus problem. I haven't found evidence yet that any
virus is involved. It looks like commercial spam, clear & simple. Our
problem is not that, however, but that they've forged addresses in our
domain in sending that spam, and that is causing us problems. There are
laws concerning fraud and denial of service. We may well have a leg to
stand upon.
I should mention that whatever Glenda and Drew are seeing is almost
certainly not related to the particular problem I'm worried about here.
|
gull
|
|
response 10 of 26:
|
Jan 12 04:31 UTC 2003 |
I'd be in favor of sending cease-and-desist letters, except that I fear that
finding a useful address to send the letters to would require lots of
expensive legal proceedings. ISPs don't turn over records casually; we'd
almost certainly have to go to court to get them before we could even send a
letter to the spammer, wouldn't we?
|
carson
|
|
response 11 of 26:
|
Jan 12 06:23 UTC 2003 |
(it's probably easier to send a letter to the ISP. that might not be
as effective, but it's something.)
|
mdw
|
|
response 12 of 26:
|
Jan 13 03:32 UTC 2003 |
That would be option #4 in the original text. I believe it will be
completely ineffective - I went through some of the IP addresses and
sites referenced in the spam, and became convinced these are
"experienced" spammers who expect to have their sites shut down
frequently. Or more precisely, I think it's one entity that does the
actual spamming, and a number of customers, who buy "advertising" from
the spammer. The "customers" would sometimes have more than one DNS
name per IP address, or the same address would appear for 2 different
company names, it all looked pretty fly by night, so I have little
confidence that complaining to the ISPs will actually solve our problem,
which is not that they send spam, but that they direct bounces at us.
|
drew
|
|
response 13 of 26:
|
Jan 13 18:30 UTC 2003 |
Re #9:
It isn't??? I rather thought that was what this item was all about. Where,
then, shall what I'm getting be discussed?
I got another one. I appended it to file spambounce.
|
mdw
|
|
response 14 of 26:
|
Jan 13 22:11 UTC 2003 |
Um, never mind - yes, drew is receiving genuine bounces from forged
spam. At least he's not receiving 6 megabytes/day, which is true for 8
other particular accounts, and what I initially feared might be the case
for drew.
|
tsty
|
|
response 15 of 26:
|
Jan 21 12:58 UTC 2003 |
is it still true that *every* isp MUST maintain a postmaster@isp-in-question
in ordre to stay nthe net?
is it true that every isp has to maintan *some* email adress hat doesn't
get filled up in order to be recognized (whateer aht means)?
|
gull
|
|
response 16 of 26:
|
Jan 21 13:56 UTC 2003 |
I don't think that's ever been true. A postmaster address is required
by one of the RFC's, I think, but it's not like there are net.cops who
will come knocking on your door if you violate that rule.
I think there is a service somewhere that runs a DNS-based blacklist of
sites that don't provide such an address, though, so people who want to
can ostracize them.
|
mdw
|
|
response 17 of 26:
|
Jan 21 19:37 UTC 2003 |
RFC 822 / 2822 require postmaster@ .
RFC 2142 requires abuse@ .
Neither of these is there per-ISP, but per "mail domain".
There are no net cops, per se. There are things one can do to piss off
important bits of the internet.
|
keesan
|
|
response 18 of 26:
|
Jan 21 22:40 UTC 2003 |
I frequently get back spam reports sent to abuse@ and then I send them to
postmaster@ and they don't come back again.
|
tonster
|
|
response 19 of 26:
|
Jan 22 07:22 UTC 2003 |
resp:18: That doesn't mean they weren't redirected to /dev/null, though.
:)
|
tsty
|
|
response 20 of 26:
|
Jan 22 10:05 UTC 2003 |
re #17 .. and those things might be ???????
|
gull
|
|
response 21 of 26:
|
Jan 22 14:00 UTC 2003 |
Re #17: Interesting. I didn't know abuse@ was required, thanks.
|
keesan
|
|
response 22 of 26:
|
Jan 22 20:29 UTC 2003 |
Today I got back another mail sent to abuse@hyomon.co.kr (or something
similar) from postmaster@ (so I sent the mail back to postmaster asking them
to close their open relay).
|
tsty
|
|
response 23 of 26:
|
Jan 24 04:42 UTC 2003 |
btw, (don't know if it's been posted before or not) yahoo has a special
adress for spam: mail-abuse@yahoo-inc.com they seem to respond beetter
at that than any other.
maybe it's better automated, but i've never beenhit twice formthe same
spammer (using or faking yahoo) when i send it to that addrs.
|
keesan
|
|
response 24 of 26:
|
Jan 24 18:48 UTC 2003 |
I get back immediate responses from sending simply to abuse@yahoo.com and
sometimes they write back saying they took action.
|