ISPs That Actively Combat SPAM?

The Llama King asks: "Like a good netizen, I do my part to report spam. But most Internet providers merely respond with a canned e-mail and it's hard to tell whether action was taken - or when. I know a lot of abuse desks are overwhelmed and spammers can get a free ride if they pick their targets carefully. Occasionally I'll get a personalized response, and even notification that a spammer's access and/or Web site has been nuked - but that's rare, and seems to be getting rarer. What ISPs are best at responding to spam complaints in a timely fashion, both in terms of killing e-mail accounts and shutting down sites that have been spamvertised?"

Comments

two of the biggest hosters/isps in germany do

[collin.m]

Schlund+Partner AG and 1&1 Internet AG, they have build a nice testing system and operate a hugh blacklist (sadly non public) here is the link

Admins seem to be lazy (slightly OT)

[Deagol]

I recently installed postfix for our domains and started rejecting IPs without a hostname (reject_unknown_client). Spam getting through dropped to a trickle, the reject-to-accept ratio being about 3:1, or about 1000 rejects a day. Unfortunately, there are many mis-configured sites out there, so some legit email was being denied.

One would think that the remote sender would complain to their mail admin first and they would get it fixed (distributed debugging, if you will). But no, they bitched to the person on my end (even though postfix's default boune messages are pretty self-evident) and then I'd end up adding an exception.

Initially, I would email {post,host}master@ the offending domain. While some were thankful for the notice, most either ignored me or flat out refused to add a rDNS entry for the mail server. Granted, it's not required by RFC, but in my opinion legit hosts should have DNS entries.

(And no, I can't just ignore the problem. When the person who writes your paycheck looses email, you're fighting a loosing battle.)

Actually, I gave up using reject_unknown_client today, except for large domains which are configured correctly (MSN, Hotmail, Microsoft, etc.) and a handful of Asian netblocks.

So back to the OP... I wouldn't hold out for admins to take care of the spam for you, especially if they're with a company you don't actually work for.

My own observations

[dacarr]

At one time, all of the now big ones - Compu$serve, Earthlink, Netcom - were very active in doing this. Mindspring was also, and I think I have a couple "we killed 'em" messages sitting around. Not anymore though - you're right, it's only canned replies as far as the eye can see.

These days if I get a response it's from Hotmail. Small ISP's also have the time for this, but small ISP's are small ISP's and tend to not require the manpower of the likes of Speakeasy, Earthlink, etc. for their basic operations - so accordingly when the occasional spammer buys usage on a small ISP, and they disuser him, they can respond to the complaints en masse and say "we got 'im, sorry 'bout that".

I think the biggest reason for this is owing to the fact that dealing with spam is unto itself a laborious task. I suppose you can set up a filter for the local abuse address to bounce around email pertaining to a specific case, but first you have to identify the case - a filter won't drop in place by itself. Then, when the problem is pinpointed to the user, you have to (in no particular order) eliminate the account (easy enough), kill the user's dialup session if necessary (why get the DSL or the T1 if you know it's going to be killed the second you start spamming?), and block his port 25 access so he can't send mail. Maybe send a little courtesy message saying "All your base are belong to us" to the spammer as you nuke his account, or set his account to download mail precisely once, and he promptly loses his connection after that. After all that's done, then you have to draft up a reply or send a canned message to the complainers.

In short, you can't win, and it sucks royally.